Journal of Geodesy, Cartography and Cadastre [615886]

Journal of Geodesy, Cartography and Cadastre

1
Tabl e of Contents

Table of Contents ………………………….. ………………………….. ………………………….. ………………………….. ……… 1

Differential SAR Interferometry for the Monitoring of the UNESCO Worl d Heritage Sites ………………………….. . 2
Iulia Dana Negula, Violeta Poenaru

Extending the availability area of positioning services ROMPOS ………………………….. ………………………….. …. 9
Mihaela -Simina Puia, Vlad Sorta, Neculai Avramiuc

Systematic reg istration of properties – a new challenge for Romanian cadastral system ………………………….. …. 15
Mușat Cosmin Constantin, Vîlceanu Clara Beatrice, Moscovici Anca Maria

Deformation prediction and analysis by finite element method ………………………….. ………………………….. …….. 21
Cătălin Io nuț Vintilă, Andreea Carmen Rădulescu, Petre Iuliu Dragomir (coordinator)

Testing a new method of digital images acquisition in the process of 3D reconstruction of an object ……………… 28
Paun Corina Daniela, Valeria Ersilia Oniga, Ciobanu Laura Elena

Investigation on the influence of the incidence angle on the reflectorless distance measurement of a terrestrial
laser scanner ………………………….. ………………………….. ………………………….. ………………………….. ……………. 37
Miriam Zámečníková, Hans Neuner, Stefan Pegritz, Robert Sonnleitner

Specifics of geodetic engineering measurements for staking out high voltage lines ………………………….. ……….. 45
Petre Dragomir, Alexandru Moldovea nu, Tudorel Silviu Clinci, Daniela Cristiana Docan

Summary of PhD thesis : Contributions in obtaining a geographic information system regarding the
archaeological area of a region ………………………….. ………………………….. ………………………….. ……………… 52
Ionut Maican

Journal of Geodesy, Cartography and Cadastre

2
Differential SAR Interferometry for the Monitoring of the UNESCO
Worl d Heritage Sites

Iulia Dana Negula, Violeta Poenaru
Rece ived: April 2015 / Accepted: September 2015 / Published : December 2015
© Revista de Geodezie, Cartografie și Cadastru / UGR

Abstrac t
Persistent Scatterer Interferometry and Small Baseline
Subset Interferometry enable the detection, measurement
and monitoring of ground displacements. These differential
syntheti c aperture radar interferometry techniques were
applied to investigate the ground stability of the Historic
Centre of Sighisoara, a cultural UNESCO World Heritage
Site. In line with the objectives of the "Open Initiative on
the Use of Space Technologies to Support the World
Heritage Convention", the monitoring of Sighisoara World
Heritage Site was performed based on Earth Observation
data. A multi -temporal series of very high resolution
TerraSAR -X data was analyzed and processed and the
results provide the local authorities in -charge of the site
protection and conservation a basis for informed corrective
measures and strategies .

Keywords
Differential SAR interferometry, TerraSAR -X, Historic
Centre of Sighisoara, World Heritage Site, Earth
Observation

1. Introduction

Earth Observation (EO) has become a valuable tool for
cultural and natural heritage monitoring considering that it
allows the generation of complete, consistent, accurate and
timely information, as a growing number of EO satellites
are launched every year. Consequently, the data archives
are continuously enlarging, thus providing a huge amount
of data that can be used to monitor the World Heritage
Sites.

Researcher I. Dana Negula; Researcher V. Poenaru
Romanian Space Agency
Address: 21 -25 Men deleev Str., 010362, Bucharest, Romania
E-mail: iulia.dana@rosa.ro; violeta.poenaru@rosa.ro Based on long -term EO data it is possible to extract useful
information that help not only to understand how a World
Heritage Site is affected by changes and the r ole played by
the human activities in these changes, but also to facilitate
the formulation and implementation of appropriate protection
and conservation policies and strategies.
One of the most important factors for sustainable strategies is
represented b y the information on ground and structural
stability. The increased development and launch of very high
resolution synthetic aperture radar (SAR) sensors opened the
door to a wide range of innovative applications. On one hand,
multi -temporal SAR satellite imagery offers a unique insight
on the past subsidence trends and provides the current status
on terrain deformation. On the other hand, very high SAR
data enable the investigation of the buildings stability. New
differential SAR interferometric (DInSAR) t echniques that
enable the ground deformation/displacement monitoring up
to a few millimeters have been developed. These techniques
are represented by Persistent Scatterers Interferometry (PS –
InSAR) and Small Baseline Subset Interferometry (SB –
InSAR) and th ey are currently the most appropriate solution
for the monitoring of a World Heritage cultural or natural
site that might be subject to ground displacement. The
displacement information represents the foundation for
sustainable heritage protection and conse rvation policies that
include the potential required mitigation measures.
UNESCO, the United Nations Educational, Scientific and
Cultural Organization, maintains the World Heritage List of
properties considered of "outstanding value to humanity" [1].
At th is moment (April 2015), the List includes 1007 World
Heritage Sites. The Romanian properties inscribed on the
World Heritage List consists of the Danube Delta, the
Churches of Moldavia, the Monastery of Horezu, the Dacian
Fortresses of the Orastie Mountain s, the Wooden Churches
of Maramures, the Villages with Fortified Churches in
Transylvania, and the Historic Centre of Sighisoara [2].
In 2003, UNESCO and the European Space Agency (ESA)
signed the "Open Initiative on the Use of Space Technologies

Journal of Geodesy, Cartography and Cadastre

3
to Suppor t the World Heritage Convention" that has the
goal to protect, monitor, document, present and share the
World Heritage sites. In this context, the term "space
technologies" refers mainly to EO and secondly to other
technologies such as navigation, position ing,
communication [3].
The potential of EO satellite imagery for the monitoring of
cultural and natural World Heritage Sites was demonstrated
by a significant number of scientific studies conducted for
various test areas [4], [5], [6], [7], [8].

2. Historic Centre of Sighisoara and very high
resolution satellite imagery

Located in the heart of Romania (N 46 13 04, E 24 47 32),
in the Mures County, the Historic Centre of Sighisoara was
inscribed on the World Heritage List in 1999 as a cultural
site due t o its architectural and urban monuments that were
built by the German craftsmen and merchants starting with
the 13th century. According to the UNESCO description [9],
Sighisoara "is a fine example of a small, fortified medieval
town which played an importa nt strategic and commercial
role on the fringes of central Europe for several centuries".
A considerable number of significant monuments and
buildings are still present today in the Sighisoara Citadel,
i.e. the Church on the Hill, the Joseph Haltrich High
School, the City Hall, the Monastery Church, the Venetian
House, the Scholars' Stairs, the Stag House, the St. Joseph
Roman -Catholic Church, the Georgius Krauss House, and
the towers that were used in the past to guard the fortified
settlement. 14 towers were originally built by the guilds of
craftsmen, out of which 9 are still standing, namely the
Clock Tower (Figure 1), the Tanners' Tower, the Tailors'
Tower, the Furriers' Tower, the Ropemakers' Tower, the
Butchers' Tower, the Blacksmith Tower, the Shoem akers'
Tower, and the Tinkers' Tower.
In the last years, the city of Sighisoara has been affected by floods that led to landslides that damaged some parts of the
Historic Centre externa l wall, thus threatening the integrity of
the World Heritage site. Such events, caused by heavy
rainfalls, have been reported in the local press and there was
a general concern that the Sighisoara Citadel might be in
danger. Moreover, the UNESCO monitoring reports [10]
corresponding to 2005 -2012 mention the "deterioration of
monuments in general and fortifications in particular" and the
"lack of protection and maintenance measures, local
responsibility and funding strategies".
In line with the objectives of the "Convention Concerning the
Protection of the World Cultural and Natural Heritage"
adopted by UNESCO in 1972, each State Party has the
obligation to evaluate the state of conservation and identify
the threats that might endanger the national World Her itage
Sites [11] and the outcomes are periodically reported to the
UNESCO World Heritage Committee.
At the moment, EO satellite imagery can be favorably used
for the monitoring of the UNESCO World Heritage Sites.
Both cultural and natural heritage can str ongly benefit from
the usage of satellite imagery, as their spatial, spectral and
temporal resolutions are constantly improving with every
new mission. In addition, the spatial resolution of the optical
and SAR imagery (especially the data acquired by the X-
band sensors) is presently adequate for the monitoring of
cultural heritage sites, considering the required level of detail
in comparison with the natural heritage sites. Apart from the
remote sensing approaches that enable only the identification
of la ndscape changes, the DInSAR techniques also provide
information on ground (and structural) displacement.
For the monitoring of the Historic Centre of Sighisoara, a
multi -temporal series of TerraSAR -X (TSX) High Resolution
SpotLight (HS) images was program med and acquired
between March and October 2014. TerraSAR -X (Figure 2) is
a side -looking X -band SAR equipped with an active antenna
which is capable of acquiring data in different imaging
modes. The platform's nominal orbit height at the equator is
514 km, orbit inclination 97.440 and the orbit repeat cycle is
11 days. TerraSAR -X was launched in 2007 and TanDEM -X
in 2010. The TanDEM -X mission has the goal of acquiring
data for a new global digital elevation model (DEM).
The HS images cover an area of appro ximately 5 km x 10 km
(Figure 3) and have 2.1 m ground range resolution and 1.1 m
azimuth resolution. Simple VV polarization and a bandwidth
of 300 MHz define the HS data. The images were acquired
from a descending orbit (orbit 47) with a view angle of 350
(beam spot 038R) at a revisiting time of 11 or 22 days. The
list of TSX HS data is presented in Table 1. The last column
of the table shows the perpendicular (normal) baselines
computed in relation to the master image acquired on the 17th
March 2014. The processing of the TSH HS data have been
performed based on ENVI SARScape, the Interferogram
Stacking module.

Fig. 1. The Clock Tower – Historic Centre of Sighisoara symbol

Journal of Geodesy, Cartography and Cadastre

4

Fig. 2. Artist impression of TerraSAR -X and its twin satellite
TanDEM -X on orbit (© Airbus Defence and Space)

Fig. 3. Te rraSAR -X preview image (medium resolution)
covering the Historic Centre of Sighisoara (© DLR 2014)
Table 1. List of TerraSAR -X HS images used for the ground and
structural monitoring of the Historic Centre of Sighisoara

TSX HS Acquisitio n date Normal baseline [24]
1. 06.03.2014 -37.036
2. 17.03.2014 0
3. 08.04.2014 -216.597
4. 19.04.2014 -70.204
5. 30.04.2014 160.622
6. 11.05.2014 23.996
7. 22.05.2014 52.552
8. 02.06.2014 -77.312
9. 13.06.2014 114.147
10. 24.06.2014 72.275
11. 16.07.2014 69.136
12. 27.07.2014 4.463
13. 18.08.2014 14.844
14. 29.08.2014 -100.386 15. 09.09.2014 -33.472
16. 20.09.2014 -117.869
17. 01.10.2014 -99.345
18. 12.10.2014 -50.991

3. Methodology and results

The monitoring of the Historic Centre of Sighisoara was
performed based on multi -temporal Synthetic Aperture Radar
(SAR) data, namely a series of 18 TerraSAR -X HS images.
SAR is an active microwave imaging system. At present,
there are numerous spaceborne SAR systems (i.e. TerraSAR –
X, TanDEM -X, Radarsat, Sentinel -1, and Cosmo -SkyMed)
that provide data for a wide range of applications [12],
including thematic mapping, topographic mapping,
deformation monitoring, and atmospheric delay mapping
[13]. A complex SAR image is a two -dimensional matrix of
pixels that store in the form of a complex number the
amplitude and the phase of the signal backscattered by the
ground targets towards the sensor [14], [15].
Geometrically similar to the conventional stereoscopic
approach [12], SAR interferometry is bas ed on the principle
that the same area on the ground is viewed from slightly
different positions [16]. This acquisition geometry can be
obtained either simultaneously (single pass / cross -track
interferometry) – with two radar systems on the same
platform – or at different moments in time (repeat pass /
along -track interferometry) – by repeated passes of the same
satellite platform [12], [17], [18]. The phase difference
between two radar images can be used to measure differences
in the geometric path length or geometric distortions in the
range direction [12], [19] . Two types of products can be
derived based on SAR interferometry, namely DEMs and
deformation/ displacement maps. In the first case the
technique is called conventional (standard) interferometry
(InSAR) and in the second case differential interferometry
(DInSAR) [20], [21]. The input data for DInSAR consists of
a pair of SAR images with a near -zero perpendicular
(normal) baseline, a pair of SAR images with a non near -zero
baseline (mandatory DEM), three SAR images, or two pairs
of SAR images [20]. In all the cases, the SAR data should be
selected according to certain criteria, out of which the value
of the perpendicular baseline is the most important [20].
Recently, new DInSAR techniques based on m ulti-temporal
(MT) SAR data have been developed [22]. According to
[22], the space -borne DInSAR approaches include PS –
InSAR, SB -InSAR, Differential SAR Tomography (D –
TomoSAR), combined MT -InSAR, and SqueeSAR. All these
techniques are suitable for cultural heritage monitoring [22].
In relation to the ground -based DInSAR approaches, Ground –
Based SAR Interferometry (GB -InSAR) and Ground -Based
Real Aperture Radar Interferometry (GB -InRAR) enable
structural deformation monitoring based on static,
respectively dy namic measurements [22].
PS-InSAR involves the identification of persistent scatterers

Journal of Geodesy, Cartography and Cadastre

5
(PSs) within the generated interferograms. PSs represent
radar reflectors that remain stable across time and with
different acquisition geometries. PS -InSAR reduces the
main errors specific to the conventional DInSAR
processing, namely temporal and geometrical decorrelation,
and atmospheric artefacts [23]. The following main steps
are performed within the PS -InSAR workflow: computation
of the complex and differential int erferograms (based on a
reference DEM), preliminary estimation of the candidate
PSs, correction of the differential interferograms, and final
estimation of the PSs in an iterative process [20], [23], [24],
[25]. In the final step, a map showing ground and structural
displacement over the selected time interval is generated by
correlating the residual phase variation (as a function of
time) to an expected temporal evolution model for each PS
[15]. This steps involves phase unwrapping and DEM and
baseline err or correction [25].
SB-InSAR uses multilooked interferograms to monitor the
evolution of ground displacement with a very high degree
of temporal and spatial coverage [26]. In the first
processing step, all the InSAR pairs with small baseline
values are se lected. Next, the corresponding interferograms
are multi -looked in order to reduce the phase noise [20] ,
[26], [27], [28] . In parallel, the coherence maps are
generated. The most coherent pixels are extracted and their
noise effects are considered negligib le [12], [26], [29]. The
processing chart continues with phase unwrapping using, in
most of the cases, the minimum cost flow (MCF) algorithm
[21], [26] . Afterward, the coherent pixels are analyzed over
the unwrapped interferograms, considering that the
topographic and atmospheric phase have been estimated
and removed [26], [29].
The DInSAR techniques have limitations and benefits. The
limitations are given mainly by the temporal decorrelation,
the atmospheric artefacts, the availability of images, and the
line-of-sight (LOS) measurement (higher performance in
case of vertical displacements) [30]. [31]. Additionally, the
magnitude of the deformation rates, the temporal and spatial
sampling of the deformation phenomena, the geocoding
precision, the linear def ormation model assumption used by
most of the processing software, and the problematic
interpretation of the deformation results might limit the PS –
InSAR relevance [31]. Nevertheless, PS -InSAR has the
great advantage of accurately (1mm/year) detecting,
measuring and monitoring ground displacements over large
areas and structural displacements of individual features
such as buildings, dams, other artificial structures [32], [31], regularly and at a relatively low cost [30]. Presently, the
enormous archives o f SAR imagery enable the investigation
of past events for which other types of data are not available
[30], [31]. In what SB -InSAR concerns, the main advantage
of the technique is that it generates deformation maps with a
high temporal sampling rate preser ving at the same time the
spatially dense displacement information [29].
The validation of the PS -InSAR results can be performed
based on leveling data collected preferably at the same time
with SAR data acquisitions [33]. Traditional leveling and
GNSS dat a provide information with low spatial sampling
density in comparison with satellite data processing [34].
Also, depending on the size of the investigated area, ground –
truth campaigns are often difficult to put in practice. It is
important to notice that t he position of the PSs cannot be
determined in advance of SAR data processing [31].
PS-InSAR have been successfully applied for the monitoring
of Sydney [25], Barcelona and other Spanish cities [30],
Amsterdam [31], Berlin [32], the German Ruhr area [33],
Tianjin suburbs of China [34], [36], Las Vegas and Paris
[35], the City of Dakar [37], Rome [38], Venice [39] and
Tuscany region [40] and the examples could continue.
Moreover, the great potential of TerraSAR -X for millimeter –
scale structural deformation w as proven by numerous
scientific studies, as presented in [32], [33], [34], [35], [36],
[39]. Thus, the significant advantages are given by the very
high spatial resolution translated into a very large number of
PSs with good long -term coherence, the less relevant
atmospheric path delay errors and the relatively short 11 -days
revisiting intervals [33], [34], [35].
The PS -InSAR displacement map (Figure 4) corresponding
to the Historic Centre of Sighisoara contains a very high
density of PSs. Only in the are a of the Citadel Hill the
number of PSs is lower due to the presence of vegetation.
There are approximately 15,000 PSs that have coherence
higher than the threshold of 0.85. The elevation information
was extracted from SRTM (© Jarvis A., H. I. Reuter, A.
Nelson, E. Guevara, 2008, Hole -filled seamless SRTM data
V4, International Centre for Tropical Agriculture, available
at http://srtm.csi.cgiar.org). Using the mean displacement
velocity of each PS, global statics have been computed. The
results show that th e mean value of the mean displacement
velocity evens a value of -0.73 mm/year (standard deviation
of ± 2.97 mm/year) with the investigated time interval
(March – October 2014). The validation of the results was
carried out based on SB -InSAR. The cross -validation showed
a high agreement between the different DInSAR techniques .

Journal of Geodesy, Cartography and Cadastre

6

Displacement
velocity
(mm/year)

Fig. 4. PS -InSAR displacement map – Historic Centre of Sighisoara (background satellite imagery © 2015 Google)

4. Conclusions

The ground a nd structural stability of the Historic Centre
of Sighisoara, a cultural World Heritage Site, was
investigated based on a multi -temporal series of
TerraSAR -X High Resolution SpotLight data that used
processed using the PS -InSAR technique. The very large
number of PSs enabled the monitoring of the ground and
structural displacements in the March – October 2014
time framework. The average overall mean displacement
velocity equalizes -0.73 mm/year with a standard
deviation of ± 2.97 mm/year.
The PS -InSAR techn ique allows the estimation of the
displacement velocity at millimeter level. Although the
technique has yet some limitations, it has the definite
benefit of providing accurate and useful displacement
information both over very large areas and at the level of
individual features, as extensively demonstrated by
previous scientific studies. By providing detailed
displacement information, the study may improve the
measures that have to be adopted by the responsible
authorities for the protection and preservatio n of the
Historic Centre of Sighisoara. The millimeter -level
displacement information provided by the space -borne
DInSAR techniques represents a preventive diagnosis
and should be validated based on ground -truth data (i.e.
leveling, GNSS or GB -InSAR/GB -InRAR data).
As future work, the development of a pilot monitoring
service for all the Romanian cultural and natural World Heritage Sites using the most appropriate Earth
Observation technology has already started. The
monitoring service will be designed to m eet the different
requirements of each site.
The monitoring of both natural and cultural sites using
advanced space technologies will offer very accurate
results that can help the local authorities to better
understand the characteristics of each site in o rder to take
the best protection and preservation measures.
Also, from the social and economic point of view, the
monitoring service will have an useful function as it
offers information regarding the local World Heritage
Sites state of conservation and s upports the
implementation of corrective measures before
irreversible damage might occur.

Acknowledgments

This work was supported by a grant of the Romanian
Ministry of Education, CNCS – UEFISCDI, project
number PN -II-RU-PD-2012 -3-0653.
The series of TerraSAR -X High Resolution SpotLight
images are courtesy of the German Aerospace Center and
the German Commission for UNESCO. The TerraSAR –
X images were received within the LAN 1988 proposal
entitled "Monitoring of Sighisoara UNESCO World
Heritage Site Us ing Space Technologies".
The authors are grateful to Emeritus Professor Dr.
Ramiro Sofronie, the Chairholder of the UNESCO Chair

Journal of Geodesy, Cartography and Cadastre

7
#177 within the University of Agronomic Sciences and
Veterinary Medicine of Bucharest, member of the
Twining and University Net working under the auspices
of the Romanian Commission for UNESCO.

References

[1]http://whc.unesco.org/en/list/ , UNESCO World
Heritage List, accessed April 2015
[2] http://www.cnr -unesco.ro/en/patrimoniu.php, Cultural
and Natural objectives from Romania included in The
World Heritage List, accessed April 2015
[3]http://www.unesco.org/science/remotesensing/?lang=e
n, Space for Heritage, accessed April 2015
[4] Wondie M., Schneider W., Melesse A., Teketay D.,
Spatial and Temporal Land Cover Changes in the
Simen Mountains National Park, a World Heritage
Site in Northwestern Ethiopia. Remote sensing, ISSN
2072 -4292, 752 -766, doi:10.3390/rs3040752, 2011
[5] Chen J., Jin S. P., Liao A. P., Zhao Y. S., Zhang H.
W., Ro ng D. W., Yang Z. J., Stereo mapping of Ming
Great Wall with remote sensing. Chinese S cience
Bulletin, vol.55, 2290 –94 doi:10.1007/s11434 -010-
4295 -9, 2010
[6] Hadjimitsis D., Agapiou A., Alexakis D., Sarris A.,
Exploring Natural and Anthropogenic Risk for
Cultural Heritage in Cyprus using Remote Sensing
and GIS. International Journal of Digital Earth vol. 6,
no.2, ISSN: 1753 -8947, doi:10.1080/17538947.2011.
602119, 2013
[7] Giri C., Pengra B., Zhu Z., Singh A., Tieszen L.,
Monitoring mangrove forest dynami cs of the
Sundarbans in Bangladesh and India using multi –
temporal satellite data from 1973 to 2000. Estuarine,
Coastal and Shelf Science, vol. 73, no.1, Elsevier
Ltd., ISSN 0272 -7714, 91 -100,
doi:10.1016/j.ecss.2006.12. 019, 2007
[8] Tapete, D, Cigna F., M asini N., Lasaponara R.,
Prospection and Monitoring of the Archaeological
Heritage of Nasca, Peru, with ENVISAT ASAR.
Archaeological Prospection, vol. 20, no. 2, Wiley
Subscription Services, ISSN 1075 -2196, 133 -147,
DOI:10. 1002/arp.1449, 2013
[9] http://w hc.unesco.org/en/list/902, UNESCO –
Historic Centre of Sighisoara, accessed April 2015
[10] http://whc.unesco.org/en/soc/?action=list&id_search
_ state=134, State of Conservation – Romanian
properties, accessed April 2015
[11] http://whc.unesco.org/en/soc/ , State of Conservation
Information System, accessed April 2015
[12] Rosen P., Hensley S., Joughin I., Li F., Madsen S.,
Rodríguez E., and Goldstein R., Synthetic Aperture
Radar Interferometry, Proceedings of the IEEE, ISSN
0018 -9219, vo l. 88, no. 3, pp 33 3-382, DOI 10.1109/
5.838084 , 2002 [13] Hanssen R., Radar Interferometry. Data
Interpretation and Error Analysis, Kluwer Academic
Publishers, ISBN 9780306476334, 0306476339, 2002
[14] Lusch, D., Introduction to Microwave Remote
Sensing, Center for Remote S ensing and Geographic
Information Science, Michigan State University,
USA, 1999
[15] Rocca F., Prati C., Monti Guarnieri A., and Ferretti
A., SAR Interferometry and Its Applications, Kluwer
Academic Publishers, Surveys in Geophysics, ISSN
0169 -3298, vol. 21, no. 2 -3, pp 159 -176, DOI
10.1023/ A:1006710731155 , 2000
[16] Crosetto M. and Perez Aragues F., Radargrammetry
and SAR Interferometry for DEM Generation:
Validation and Data Fusion, Proceedings of SAR
Workshop: CEOS Committee on Earth Observation
Satelli tes, ISBN: 9290926414, Toulouse, France,
1999
[17] Richards M., A Beginner's Guide to Interferometric
SAR Concepts and Signal Processing, IEEE A&E
Systems Magazine, vol. 22, ISBN 0018 -9251, 2007
[18] Massonnet D. and Souyris J. C., Imaging with
Synthetic Aperture Radar, CRC Press, Taylor &
Francis Group, ISBN 978 -0-8493 -8239 -0 (CRC
Press), USA, 2008
[19] Zhou X., Chang N. B. and Li S., Applications of
SAR Interferometry in Earth and Environmental
Science Research, Sensors, ISSN 1424 -8220, no. 9,
1876 -1912 , 2009, DOI: 10.3390/s90301876
[20] Ferretti A., Monti Guarnieri A., Prati C., Rocca F.
and Massonnet D., InSAR Principles: Guidelines for
SAR Interferometry Processing and Interpretation,
ESA Publications, ESTEC, ISBN 92 -9092 -233-8,
ISSN 1013 -7076, Noord wijk, Netherlands , 2007
[21] Tran V. A., Masumoto S., Raghavan V. and Shiono,
K., Accuracy of Low Relief Topographical Map
Derived from JERS -1 SAR Interferometry in Hanoi,
Journal of Geosciences, volume 50, art. 8, Osaka,
Japan, 2007
[22] Zhou W., Chen F., Guo H., Differential Radar
Interferometry for Structural and Ground
Deformation Monitoring: A New Tool for the
Conservation and Sustainability of Cultural Heritage
Sites. Sustainability (Basel, Switzerland), ISSN 2071 –
1050, 1712 -1729, doi:10.3390/su702171 2, 2015
[23] Yu J., Hay -Man A., Jung S., Ge L. and Rizos C.,
Urban Monitoring Using Persistent Scatterers InSAR
and Photogrammetry, Proceedings of the XXI ISPRS
Congress, Commission IV, volume XXXVII, Beijing,
China, 2008
[24] Colesanti C., Ferretti A., N ovali F., Prati C. and
Rocca F., SAR Monitoring of Progressive and
Seasonal Ground Deformation Using the Permanent
Scatterers Technique, Geoscience and Remote
Sensing, IEEE Transactions on Volume 41, 2003

Journal of Geodesy, Cartography and Cadastre

8
[25] Yu J., Ng A., Jung S., Ge L., and Rizos , C., U rban
Monitoring Using Persistent Scatterer InSAR and
Photogrammetry, The International Archives of the
Photogrammetry, Remote Sensing and Spatial
Information Sciences, vol. XXXVII, part B1. Beijing,
257-261, 2008
[26] Tizzani P., Berardino P., Casu F., Eui llades P.,
Manzo M., Ricciardi G.P., Zeni G. and Lanari R. ,
Surface deformation of Long Valley caldera and
Mono Basin, California, investigated with the SBAS –
InSAR approach, Science Direct, Elsevier, Remote
Sensing of Environment 108 , p. 277 -289, 2007
[27] Rabus B., Eineder M., Roth A. and Bamler R., The
Shuttle Radar Topography Mission – A New Class of
Digital Elevation Models Acquired by Spaceborne
Radar, ISPRS Journal of Photogrammetry & Remote
Sensing, volume 57 , 2003
[28] Manjunath D., Earthquake Inte raction Along the
Sultandagi -Aksehir Fault Based on InSAR and
Coulomb Stress Modeling, Dissertation, University of
Missouri, Columbia, USA, 2008
[29] Casu F., Manzo M., Berardino P., Fornaro G., The
SBAS approach for Earth Surface Deformation
Analysis, IEE E GOLD Remote Sensing Conference,
Italy, 2004
[30] Tomas R., Romero R., Mulas J., Marturia J.,
Mallorqui J., Lopez -Sanchez J. M., Herrera G.,
Gutierrez F., Gonzalez P. J., Fernandez J., Duque S.,
Concha -Dimas A., Cocksley G., Castaneda C.,
Carrasco D., Bla nco P., Radar interferometry
techniques for the study of ground subsidence
phenomena: a review of practical issues through cases
in Spain, Environmental Earth Science, 163 –181,
DOI 10.1007/s12665 -013-2422 -z, 2014
[31] Crosetto M., Monserrat O., Jungner A. and Crippa
B., Persistent Scatterer Interferometry: Potential and
Limits, Proceedings of ISPRS Workshop, High –
Resolution Earth Imaging for Geospatial Information,
Hannover, Germany, 2009
[32] Gernhardt S., Adam N., Eineder M. and Bamlera R.,
Potential of V ery High Resolution SAR for Persistent
Scatterer Interferometry in Urban Areas, Annals of
GIS, 16:2, DOI:10.1080/19475683.2010.492126,
2010 [33] Wegmuller U., Walter D., Spreckels V., Werner C.,
Nonuniform Ground Motion Monitoring with
TerraSAR -X Persiste nt Scatterer Interferometry,
IEEE Transactions on Geoscience and Remote
Sensing, vol. 48, no. 2, 2010
[34] Luo Q., Perissin D., Lin H., Zhang Y. and Wang W.,
Subsidence Monitoring of Tianjin Suburbs by
TerraSAR -X Persistent Scatterers Interferometry,
IEEE Journal of Selected Topics in Applied Earth
Observations and Remote Sensing, vol. 7, no. 5, 2014
[35] Eineder M., Adam N., Bamler R., Yague -Martinez
N. and Breit H., Spaceborne Spotlight SAR
Interferometry with TerraSAR -X, IEEE Transactions
on Geoscience a nd Remote Sensing, vol. 47, no. 5,
2009
[36] Yu B., Liu G., Zhang R., Jia H., Li T., Wang X., Dai
K., Ma D., Monitoring subsidence rates along road
network by persistent scatterer SAR interferometry
with high -resolution TerraSAR -X imagery, J. Mod.
Transpor t., 21(4):236 –246, DOI 10.1007/s40534 -013-
0030 -y, 2013
[37] Cozannet G., Raucoules D., Wöppelmann G.,
Michele M., Poupardin A., InSAR Monitoring of
Ground Motions Impacts for In -Situ Sea Level
Measurement: the Example of Dakar (Senegal),
IEEE, Proceeding o f IGARSS 2014, 978-1-4799 –
5775 -0/14, 2014
[38] Cigna F., Lasaponara R., Masini N., Milillo P.,
Tapete D., Persistent Scatterer Interferometry
Processing of COSMO -SkyMed StripMap HIMAGE
Time Series to Depict Deformation of the Historic
Centre of Rome, Italy . Remote sensing (Basel,
Switzerland), ISSN 2072 -4292, 12593 -12618,
doi:10.3390/rs61212 593, 2014
[39] Strozzi T., Tosi L., Teatini P. and Wegmüller U.,
Monitoring Land Subsidence in the Venice Lagoon
with TerraSAR -X, TerraSAR -X Science Meeting,
2008
[40] Tapete D and Cigna F., Rapid Mapping and
Deformation Analysis over Cultural Heritage and
Rural Sites Based on Persistent Scatterer
Interferometry. International Journal of Geophysics,
Hindawi Publishing Corporation, 1 -19

Journal of Geodesy, Cartography and Cadastre

9
Extending the availability area of positioning services ROMPOS

Mihaela -Simina Puia, Vlad Sorta, Neculai Avramiuc
Rece ived: April 2015 / Accepted: September 2015 / Published : December 2015
© Revista de Geodezie, Cartografie și Cadast ru/ UGR

Abstrac t
In Romania there are installed 74 GNSS permanent
stations, since 1999 and in 2012 the network was
completed, touching the density recommended by
European organizations. This fact allows providing high
quality ROMPOS (Romanian Position Determination
System) services inside the network. Providing reliable
ROMPOS services in the frontier zone is possible due to
cross – border GNSS raw data exchange from close
border GNSS reference stations. Nowadays Romania,
through National Agency for Cad aster and Land Registry
has signed a few cross – border GNSS data agreements
with homologous authorities from Hungary, Moldova
and Ukraine and has good cooperation perspectives by
similar authorities from Serbia and Bulgaria, with whom
these kind of data e xchanges were already tested.
ROMPOS real time services are accessed through
internet data services. Considering the fact that there are
a few areas in Romania where these kind of data services
have no coverage, to use the real time positioning
services i t is required to use alternative methods to
retransmit the products afferent to these services.

Key words
Permanent station, position determination service,
RTK/DGNSS, user

Doctoring Eng., M. S. Puia
National Agency for Cadaster and Land Registration – Cadastre
and Geodesy Department
Address No. 202A Splaiul Independentei, 1st floor, sector 6,
Bucharest, Romania
E-mail: mihaela.puia@ancpi.ro
PhD. Eng., V. Sorta
National Agency for Cadaster and Land Regist ration – Cadastre
and Geodesy Department
Address No. 202A Splaiul Independentei, 1st floor, sector 6,
Bucharest, Romania
E-mail: vlad.sorta@ancpi.ro
PhD. Eng., N. Avramiuc
E-mail: neculai.avramiuc@gmai.l.com 1. Introduction

Transition from GNSS post processing methods to real
time determination requires the achievement of regional,
national and local GNSS augmentation systems.
These systems provides to users aux iliary information,
called differential corrections for compassing centimeter
level in precision of real time positioning. Such kind of
system, including these services is also Romanian
Position Determination System – ROMPOS, realized by
National Agency f or Cadaster and Land Registry
(NACLR), according to standards elaborated by
European organization EUPOS (European Position
Determination System) for realizing in Central and
Eastern Europe an integrated system for such kind of
services. EUPOS services are designed for provide an
homogenous positioning in terms of performances, using
DGNSS and RTK methods, this being guaranteed inside
of EUPOS coverage area. For touching this goal, a good
cooperation between neighbors and members is required.
This paper pres ents and argues the necessity of GNSS
data exchange by all Romania`s neighbor countries.
ROMPOS real time services are provided only via
internet and are accessible only in areas covered by
mobile data services. For areas where internet is
unavailable are usable alternative methods to retransmit
differential corrections and these methods will be also
presented in this paper.

2. Regional and local GNSS permanent
networks

a. EUPOS network
EUPOS is a regional homogenous GNSS network ,
realized by 18 specialize d institutions in Central and
Eastern Europe, adopting some common standards,
normative, methodologies and procedures.

Journal of Geodesy, Cartography and Cadastre
10
Tabel 1: Numărul stațiilor ROMPOS incluse în EUPOS

EUPOS
National
system Area
(km2) Planned
reference
stations Realized EUPOS
referenc e stations
ROMPOS 237.500 73 74

Fig. 1: EUPOS reference stations – 459 stations [1]

b. ROMPOS network
Nowadays, National GNSS Permanent Network includes
74 reference stations (Fig. 3) , coherently and uniformly
distributed across the country. After Se ptember 2008,
this network became the framework of ROMPOS,
providing for the first time in Romania real time precise
positioning services.
In terms of equipment, the network is inhomogeneous.
There are two manufacturers: Leica and Topcon (F ig. 2).

Fig. 2: Types of ROMPOS stations

Starting with NACLR participation in the frame of
EUPOS project, was taken into account the compliance
of specific products and services standards (EUPOS
Technical Standards, Guidelines For Single Site Design,
Guideline for Cross -Border Data Exchange, Guideline
for EUPOS Reference Frame Fixing ).

Fig. 3: National GNSS Permanent Network [2]

Quality of EUPOS/ROMPOS services depends more by
choice of installation place and design.
EUPOS member coun tries must follow the installation
standards to guarantee the national networks high quality.
Standards compliance lead to homogenous services
supply in the entire EUPOS coverage area, unconcerned
the user position.
National Center for ROMPOS Services (NCR S) is a part
of NACLR and it works based on EUPOS standards,
implementing all requirements included.

3. General features of ROMPOS

ROMPOS is based on a framework of GNSS reference
stations installed and managed by NACLR. There is a
single exception the IGS station from Bucharest, called
BUCU, which is installed and managed by Technical
University of Civil Engineering of Bucharest through
Faculty of Geodesy.
GNSS reference stations works permanently, providing
real time data and post processing data in sessio ns of 1h
and 24h. GNSS reference stations are interconnected,
including the neighbor stations (near the border area).
The average distance between stations is about 70 km.
The equipment were installed to provide the long term
stability of the GNSS antennas . These antennas are
placed to ensure a “free obstruction” skyline and to avoid
the sources of bias and multipath. Using right calibrated
antennas, the multipath effect can be reduced. Most of
the antennas of ROMPOS stations were individual
absolute calibr ated, using the best techniques available in
the world. Also, the receivers are double frequency and
geodetic class.
The stations receive signals from NAVSTAR GPS
system and 69 are GPS – GLONASS. When GALILEO

Journal of Geodesy, Cartography and Cadastre
11
will be in Full Operational Capability (FOC), w ill be
proposed an upgrade for equipment.
Coordinates of the GNSS antennas are computed with a
high precision (better than 1cm) in ETRS89 (European
Terrestrial Reference System 1989) , densifying the
EUREF stations (București, Bacău, Baia Mare,
Constanța, D eva).
There are also a quality management system to support
users to obtain best results in terms of precision, integrity
and system availability. It is important to ensure a high
level of availability (minimum 99%) and integrity of the
entire system. The malfunctions, interruptions and
quality decrease of data are identified real time by the
administrators.
The national reference stations are compatible with the
most of other GNSS systems, ROMPOS being
interoperable with EUPOS.
ROMPOS services are provided by NACLR, through
National Center of ROMPOS Services, which is intended
to monitor and control the GNSS stations activity for the
automatic transfer of data to central data server, which
are used for post processing position determination (the
ROMPOS – GEO service) and to provide positioning
services and products for real time applications.

Fig. 4: National Center for ROMPOS Services (NCRS)

NCRS structure (Fig. 4) recommended by EUPOS
include:
– national GNSS reference netwo rk;
– the center for data processing ;
– the system administrators. Each GNSS station has also
its own administrator. CNSR is also supported by IT
specialists from the IT department of the institution;
– for a continuous improvement of the ROMPOS
services, NCRS cooperates with other national and
international organisations;
– NCRS directly cooperates with similar EUPOS centers,
especially from neighbor countries; – NCRS specialists regularly participate to EUPOS
Council meetings and has suffrage in this stru cture;
– now the ROMPOS real time services are provided free
of charge to a wide range of users.

4. Services offered by ROMPOS

ROMPOS includes 3 main types of services:
– ROMPOS DGNSS – for real time kinematic
applications, with an accuracy range from 3m t o 0,5m;
– ROMPOS RTK – for real time kinematic applications,
with an accuracy range from 0,5m to 2cm;
– ROMPOS GEO (Geodet ic) for post processing
applications, with an accuracy better than 2cm.
The transfer of differential corrections from reference
stations to user may be realized as follows: radio waves,
mobile communication systems GSM/GPRS or internet.
There are also, two main types of real time positioning
products: single base (from one site) and network, such
as VRS, FKP și MAC (MAX și i -MAX).
Transmission of OMPOS RTK and DGNSS differential
corrections is centralized, from NCRS via internet. The
RTK receiver must be equipped with GSM modems to
access internet and receive differential corrections. In this
case, wi th one single GNSS receiver, real time users can
be determine very quickly (seconds or minutes) their own
position and also can check the precision of positioning.
Coordinates are determined relatively to A class national
geodetic network and due to this f act this values are
obtained directly in European Terrestrial Reference
System 1989 (ETRS89) which offers an homogeneity of
all determinations.
To transform coordinate values from ETRS89 to
National Reference System S42 (ellipsoid Krasowsky
1940, Projectio n Stereographic 1970), NACLR offers for
free the TransDatRO software, which can be download
from official website: www.ancpi.ro .
For users who want the implementation of TransDatRO
software in GNSS RTK field receivers, NACLR provide
the grids for transformation and the computing algorithm.
As a result, coordinates determined by GNSS RTK field
receivers which have implemented the transformation
grids of TransDatRO application can be obtained directly
in the field (on the whole country) in S42 system.
NACLR also realized the web site www.rompos.ro
which deliver information about the system.
ROMPOS is managed through a specialized software
package for post processing and real time service s, Leica
GNSS Spider. This software packages realizes:
monitoring and control of GNSS reference stations,
managing and archiving the post processing data,
qualitative and quantitative data analyses, network
configuration, creation, management and delivery of
specific products and services, monitoring of the system
users etc. Also, system is highly secure, computation and
network management being separated from data delivery,

Journal of Geodesy, Cartography and Cadastre
12
thus being realized the infrastructure and user data
protection.
Due to the large d imensions of the network which
implies a lot of calculations, the network is divided in 4
clusters which run on two servers. This architecture
realize the processing on two servers and, in case of
malfunction not being affected the whole network. The
clust ers have common stations between them, which
permits their superposition (see Fig. no 5).

Fig. 5: The clusters of ROMPOS network

5. Means to extend the ROMPOS real time
services availability

Considering only GNSS reference stations of each
country, accuracy and reliability of positioning services
will be degrade in border zones due to mission of some
geometric information outside the network.
The exchange of raw GNSS data delivered by near
border GNSS permanent stations may be helpful for
every EUPOS National Services Center for guaranteeing
homogenous accuracy and performances of the system.
On this line, EUPOS Technical Working Group realized
a document „Guideline for Cross -Border Data Exchange ”
which regulates the general frame for EUPOS sta tions for
choosen the data exchange, data format and transport
protocol.
GNSS reference networks are used to generate
differential corrections, due to fact that is possible to
model and correct the distance errors, which reduce the
RTK and DGNSS precision of positioning commensurate
to distance between rover and GNSS reference station.
The most important errors which affect the GNSS
positioning are caused by ionosphere, troposphere and
satellite orbits. The goal of GNSS network is to model
and estimate thes e errors and to deliver differential
corrections to the real time users for obtaining better precisions in position determinations.
EUPOS reference stations have different receiver types.
Also, different reference stations and software
applications are use d to offer standard EUPOS services.
For this reason it is important to ensure GNSS real time
data exchange, in standard formats. EUPOS members
uses RTCM (Radio Technical Commission for Maritime
Services) standards for cross border data exchange. As a
resul t of RTCM standards evolution, the content and
format of the GNSS data stream are improved
significantly, message types being improved. It is
recommended to exchange data in RTCM 3.x format for
EUPOS data centers, according to [3].

Table 2: GNSS data exc hange agreements between Romania
and neighbor countries
Agreement GNSS data stations which are the
subject of agreement Signed
Romania Neighbor countries
ROMPOS –
GNSS Net.hu
(Hungary ) Arad
Oradea
Satu Mare
Timișoara Vásárosnamény
Debrecen
Gyula
Szeged Yes
2010
ROMPOS –
MOLDPOS
(Moldova ) Dorohoi
Flămânzi
Iași
Vaslui
Târgu Bujor
Brăila Cahul
Edineț
Falești
Nisporeni
Leova
Comrat Yes
2010
ROMPOS –
ZAKPOS
(Ukraine ) Satu Mare
Baia Mare
Vișeu de Sus
Dorohoi
Tulcea Khust
Rakhiv
Chernovtsi
Izmail Yes
2012
ROMPOS –
BULiPOS
(Bulgaria) 5 stations 5 stations No
ROMPOS –
AGROS (Serbia) 4 stations 4 stations No

GNSS data streams received from neighbor stations are
used only for ROMPOS network products, for extending
the coverage area of network services, according to cross
border GNSS data exchange agreements. RINEX files
should not be generated from public data streams.
For out of network areas for which the GNSS data
exchange is realized, there is modeled the errors due to
ionosphere, troposphe re and satellite orbits, and
corrections delivered to users are reliable. These real
time services offered by extended systems are quality,
being able to obtain precise positioning (see Fig. 6).

Journal of Geodesy, Cartography and Cadastre
13

Fig. 6: Outside network measurements , near Hungary and Uk raine
borders – user of Ro – MAC product which has “fixed” type
solution

On the other side, in areas where the cross border data
exchange is not realized (Serbia and Bulgaria), precise
position determination of the users is more difficult (see
Fig. 7).

Fig. 7: Measurements outside network , where the data exchange is
not realized (Serbian and Bulgarian borders) – network products
users (real time RTK ) which have “float” type solutions

Delivery of real time services from positioning services
center to the users is realized via internet, using
GSM/GPRS connections. Thereby access to rhis services
implies data services coverage.
In areas where internet access is unavailable (due to lack
of GSM/GPRS data coverage) and there are necessary
real time servic es, products of these services can be
transmitted through a radio transmitter. In these
conditions are used auxiliary equipment which takes
differential corrections transmitted by positioning
systems and retransmit them, via radio waves, where the
data sig nal coverage is unavailable.
This system should have a good operating autonomy, regardless of field conditions, thus being possible by
using auxiliary power supplies, protection against
damage , transport etc. Through this kind of equipment
the coverage are a of real time services will be extend.
There are many cases of using these devices :
1. Radio transmitters have an important usefulness where
the data signal coverage doesn’t exists and through these,
the availability area of differential corrections can b e
extended. In this case the radio transmitter is installed in
a covered area with data signal, which allows the
connection to GNSS network server. Thereby,
differential corrections are taken via internet and
retransmitted in the uncovered interest area wh ere users
equipped with RTK receivers works. Thus, the correction
message can be accessed by an unlimited number of
receivers equipped with an adequate radio modem. For
realizing good quality determinations using differential
corrections, the approximate p osition which will be
transmitted to server will be from work area and not the
position of the radio transmitter, due to fact that distance
between them may be long.
2. Radio transmitters may be used also in areas where the
mobile data coverage exists, but it whishes that field
receivers to receive a specific message, which the
network server don’t provide. In this situation, the system
which contains the radio transmitter in the work area will
transmit to the NTRIP server an approximate position,
for which will receive differential corrections. These
corrections are received through NTRIP server via
internet and will be retransmitted to the field rovers
through a radio modem in a specific data format, which
can be accessed by receivers.

Fig. 8: The exte nsion of differential correction transmission using a
radio transmitter and supplementary radio stations

3. For transmitting the data stream using much frequency
or much differential corrections message types, should be
connect another radio transmitter – a supplementary
radio base (see Fig. 8) . Thereby the use of differential
corrections to different rover models is improved.
The RTK data are delivered via mobile internet in RTCM

Journal of Geodesy, Cartography and Cadastre
14
SC 104 vs. 2.x și 3.x standard data formats. In case of
using a proprietary data format, the GNSS receiver must
be able to decode the data streams in such way that
software installed on field equipment to be able to
process it.
The National Center f or Cartography have two similar
systems for retransmission of differential corrections,
which are equipped with all tools necessary for the field
work whereby it is possible to work in uncovered data
signal areas.

6. Conclusions

Position determination serv ices are designed to ensure a
consistent positioning in term s of performance, using
DGNSS and RTK methods , this being guaranteed
throughout the coverage area of the network underlying
the system. In order to achieve this objective it is
necessary the coope ration between neighboring
networks. Positioning accuracy and reliability of services
will degrade near border areas due to the omission of
information from some geometric information . By
signin g agreements on the exchange data from similar
permanent netwo rks from Bulgaria and Serbia, will pave
the provision of real -time quality data also in the border
areas.
Accessing services created on the basis of permanent
station networks should be as easy user to cover as many
of these requirements. This can be achie ved through the
use of services and equipment to complete th em.

References

[1] http://www.eupos.hu/map_out.php , viewed at
02/06/2015
[2] www.rompos.ro , viewed at 02/06/2015
[3] www.eupos.org – Guideline for Cross – Border Data
Exchange
[4] Cooperation agreement between the Institute of
Geodesy, Cartography and Remote Sensing of
Hungary and the National Agency for Cadastre and
Land Regis tration of Romania regarding GNSS
reference station data echange
[5] Cooperation agreement regarding GNSS data
echange between the National Agency for Cadastre
and Land Registration of Romania and the Agency
for Land Relations and Cadastre of Moldova
Republic
[6] Cooperation agreement between the National
Agency for Cadastre and Land Registration of
Romania and State Service of Geodesy, Cartography
and the Cadastre Ukraine/ State Enterprise
«Zakarpatgeodezcentre», regarding GNSS reference
station data exchange

Journal of Geodesy, Cartography and Cadastre
15

Systematic registration of properties – a new chall enge for Romanian
cadastral system

Mușat Cosmin Constantin, Vîlceanu Clara Beatrice, Moscovici Anca Maria
Rece ived: April 2015 / Accepted: Septemb er 2015 / Published : December 2015
© Revista de Geodezie, Cartografie și Cadastru/ UGR

Abstrac t
The dynamic international professional environment and
the constant contacts among similar institutions have
contributed in time to the c reation of a common vision on
relevant issues and identification of common targets,
meant to diversify the products and satisfy ever more
customers and make them loyal; . that is precisely why we
consider not only important, but also interesting that one
should learn as many things as possible about the activity
performed by National Agency of Cadastre and Land
Registration in the field of international relations, within
a multicultural professional environment in which, due to
common efforts, we have our pa rt, very well defined.
For many decades, traditional cadastral systems have
tended to enjoy a reputation for reliability, well defined
processes, and a well -recognized guarantee of security of
private land ownership. Tremendous technological
progress, soci al change, globalization, and the increasing
interconnection of business relations with their legal and
environmental consequences, however, have put a strain
on the traditional systems.
They cannot adapt to all the new developments. An
obvious indication of this is the many reforms that
cadastral systems are going through.

Keywords
Systematic cadastre, land registration, real estate,
cadastral system, land ownership.

Lecturer Ph.D. Mușat Cosmin Constantin
Assitant Ph.D. Vîlceanu Clara Beatrice
Ph.D . Student Moscovici Anca Maria
Politehnica University of Timișoara, Faculty of Civil Engineering
Timișoara, 2A Lalescu Street, room C209
E-mail: cosmin.musat@upt.ro ; beatrice.vilceanu@upt.ro ;
moscovicianca@yahoo.com
1. Introduction

Nowadays, the Romanian cadastral system faced a new
provocation, namely to elaborate and sustain a new
approach of level registration, which includes parcels,
buildings and legal owner. Today, there are four main
directions of cadastral activities in Romania:
– moving towards a modern Land Administration –
Transition from Sporadic to Systematic registration;
– the integrated Electronic Cadast re and Land Registry
in Romania using an informational interface called E –
TERRA;
– CESAR project financed by World Bank for
continuo s systematic registration of all properties;
– Romanian Government and National Agency of
Cadastre and Land Registration project for an estimated
40 million estates in all counties between 2015 -2023.

Fig. 1. Data model approach for easy and fast registration [10]

The major achievements of these challenges could be
classified as follows:
– establishing unified institution for cadastre and land

Journal of Geodesy, Cartography and Cadastre
16
administration;
– full operational centralized IT system managing
sporadic registration;
– start of systematic registration projects from both loca l
and central level with World Bank financial support;
– political and social awareness of limitations of the
current system. [4]

After the radical changes in the former communist
countries, the cadastre has gained importance to secu re
the newly created private land rights and to support the
emerging land markets in the transition nations. But also
in all countries where private land ownership or even
land use rights of the citizens exist, cadastre is an
important issue (Fig. 1). With out a reliable
documentation on the legal situation concerning land,
economic development is hampered. [5]
Cadastres (Fig. 2) together with the land registries help to
transform the land rights and resources into tradable
assets.
Cadastres and land registr ies document and localize
pieces of land, which are objects of ownership or land
use rights, which are defined in the constitutions and
laws in many countries of this world. [3]

Fig. 2. Correlation between documents for a modern cadastral
concept

Realizing cadastre is a process which involves all legal
owners from an administrative territorial unit. This fact
implies effective partic ipation of the legal owners by
their engagement to field activities and publishing the
results .

2. Major problems of actual cadastral system

According to Romanian Law on cadastre and land
registration no. 7/1996, general cadastre is the unitary
and compuls ory system of technical, economic and
juridical inventory of registering of the whole land fund of the country (Fig. 3), regardless of the destination of
land or related constructions, or the quality of owners.
[1].

Fig. 3. Examp le of land fragmentation (Romanian village)

The basic entities of this system are the parcel, the
building and the owner. Immovable means one or more
adjacent parcels, with or without buildings, belonging to
the same owner. Parcel means the land area havi ng the
same land use. Cadastre and Land Registration Law no.
7/1996 laid the legal foundations of a unique system of
land registration for the entire country, which was to
replace the existing land registration systems: the one of
the registers of transcri ptions and inscriptions, regulated
by Civil Code and Civil Procedure Code, the one of the
former land books respectively, regulated by Decree
no.115/1938 regarding unification of provisions on land
books. As its name suggests, the new law regulates two
different institutions in the same normative instrument:
the technical institution of the cadastre and the legal
institution of new land books.
The cadastral and land registration entry sys tem keeps
records of: immovable , property rights, inscriptions on
divisions of property rights and charges (Fig. 4).
Buildings represent a special case of registration if their
owner is different from the one of the land, as well as
apartments for the blocks of flats. The buildings on the
parcel are registered together with this if the owner is the
same.

Fig. 4. Inconsistencies of land application laws

Journal of Geodesy, Cartography and Cadastre
17

Fig. 5. Sporadic cadastre map

The amount of legal procedure has caused through the
years, multiple issues in the field work s, which became
major problems of actual cadastral system (Fig. 5):
– land fragmentation (land reform / land management
projects before the reform);
– inconsistent land information as a result of land
reform
(lack of modern standards when land reform was star ted);
– large areas with no cadastral information;
– limitation of the current sporadic registration
procedure with risks on secure system. [6] .

3. Technical infrastructure of the E -TERRA
cadastral system and e -CF – the Romanian
real estate land book

The inte grated cadastre and land registration IT system
(e-Terra) is National Agency of Cadastre and Land
Registration’s main operational computer system that
ensures the management of the electronic cadastre and
land registration records.
The purpose of this comp uter system is to unify,
standardize and automate the processes of updating and
inquiring the cadastral and juridical records administered
by NACLR, having as final target the increase of quality
of services delivered to the citizens and the institutio ns of
this country.
E-Terra system functional components (also called
modules or applications) are divided into main
components and auxiliary components depending on
their contribution to the achievement of the general
purpose of the system: cadastral and legal records
management.
Each module (Fig. 6 ) offers functionalities for the
settlement of specific problems related to the cadastral
and legal records management activity described
hereinafter.

Fig. 6. The Operational System of e -Terra

The e- C(arte)F(unciara ) system ensures the electronic
filling in and solving of the applications made for land
book and cadastre services provision by professional
groups (notaries, banks, real estate agencies, licensed
persons etc.) (Fig. 7 ). The functionalities offer ed by the
system are: solving of applications for information
extract, authentication extract, inscription in the land
book of the applications for the inscription in the
registers of transcriptions and inscriptions, user
authentication based on digital ce rtificate, signature of
documents and applications enclosed by professional
groups having qualified digital certificate, remittance slip
for notaries, payment modalities (advance payment,
electronic payment, regular payment).
For every real estate, a diffe rent land book is drawn up,
in which juridical acts and deeds regarding that
immovable are registered based on the application
submitted by applicants. The land book represents the
basic unit of the legal records.

Fig. 7. Users of e -Terra and e -CF

The process of real estate registration has been
automatized. [7] Thus, information is registered in a
digital database which contains both graphical (real
estate boundaries, buildings boundaries) and textual data
(cadastral number, area, owner’s name, real es tate and
owner’s address, supporting deeds of the property right
over the real estate). NACLR also owns the digital

Journal of Geodesy, Cartography and Cadastre
18
orthophotomaps, on a 1:5000 scale, for the whole
country, used both for cadastral documentations
acceptance and for preparation of projects by the
persons/institutions concerned. [2]
The use of e -Terra integrated cadastre and land
registration computer system has started as of June 2008.
Currently, the system operates in all 42 counties, and has
as main objectives: standardization of processes related
to document verification, reception and inscription in the
land book, computerization of main NACLR/OCLR
workflows, automation of certain processes and
achieving of operations efficiency.

4. The Implementation Phases of the
CESAR Project in Romania

Making the general cadastre is of interest to all the
building owners in a territorial administrative unit. This
implies their effective participation to the process, by
involving themselves not only in the field activities but
also in the stage of the p ublication of results.
In the process of general cadastre works juridical
information will be used, i.e.: the ones which envisage
the identification of all the owners, the transcript of the
property right in the land registry, as well as the
transcript of other real and possession rights. This will be
done according to transfer acts, constitutive or
declarative rights, as well as on the grounds of possession
documents which prove the possession ex pert rights.
The CESAR Project (Complementing EU Support for
Agricultural Restructuring) envisages the inventory of all
real estate and the free registration of real estate rights in
91 territorial administrative units in the rural areas. The
project is worth EUR 51.4 million and its was scheduled
completion date until June 2013 . Its results will have
profound implications for the economy, setting a solid
foundation for the development of national infrastructure
and agriculture in the geographical areas covered by the
project.
The project envisages the free cadastral and land roll
registration of all real estate, by identifying the legal
owners. The first two phases encompass systematic
registration services for approximately 423,000 ha,
respectively 620,000 properties.
Phase 1 – pilot: 19 Territorial administrative u nits, with
an estimated 225,000 buildings on an area of
approximately 150,000 ha;
Phase 2 – 31 Territoria l administrative units, with an
estimated 395,000 buildings on an area of approximately
273,000 ha.
Phase 3 – 41 Territorial administrative units, with an
estimated 402,000buildings on an area of approximately
302,000 ha.

Fig.8 – The Pilot Project – Implementation of CESAR

Table 1 – The fi rst 6th completed administrative territorial units –
Phase I of the CESAR project

Phase I and II are in progress, and phase III shall start
only after the approval for extending the project with 2
years. NACLR takes all the necessary measures to extend
the CESAR project or to obtain funds from the state
budget in order to continue the systematic registration of
properties.
This project envisages the free cadastral and land roll
registration of all real estate, by identifying the legal
owners.

Fig.9 – The second phase of CESAR Implementation

Journal of Geodesy, Cartography and Cadastre
19
5. Systematic registration of properties –
The future of Romanian Cadastral system

Providing free sy stematic cadastral and land roll
registration services (systematic recording) envisages
three main phases:
– the phase for performing preliminary works –
establish
a data base in electronic format, through: digital
conversion of existi ng analog archives (lan d rolls,
carto graphic documents, scanning and georefe rencing
maps and plans scale 1: 10.000 – 1:500), carry out
orthophotoplans;
– the execution phase of the cadastral plan and registry .
It shall contain:
– the limits of territorial administrative units;
– the limits of real estate (lands and buildings) inside
– the built -up area, as well as outside the built -up area;
– the cadastral numbers of all real estate inside and outside
the build -up area.

Table 2 – Required funds for applying the project

– the land roll ex officio registration phase. The opening
of land rolls will finalize the identification of limits and
owners of real estate, existing at the time of the
systematic registration works. All the subsequent
changes, referring to the area and owners of real estate,
shall be registered in the cadastral and real estate
publicity records based on maintenance works of the
cadastre.

The program is performed:
– based on the strategy of the National Agency for
Cadastral Surveys and Real Estate Publicity, approve d by
the administration committee;
– with expenses financed by:
– state budget;
– budget of territorial administrative units;
– public -private partnership – concession of services
– international sources.
A reliable real estate publicity system would ensure the
ownership right through precise recordings, thus contributing to the fast and safe development of
numerous national infrastructure elements, and
eventually, it would function as a prerequisite for
national prosperity.
The economical quantification of t he Project is based on
the statistical data of ANCPI, as well as on the status and
results of the projects carried out so far.

Table 3 – Areas and Territorial administrative units

The systematic registration will be developed from 3
levels:

First level – Public auction for 10 lots with 5 -15
administrative territorial units (UAT)/lots from 41
counties:
– 10 lots with 147 UAT consisting 853.458 ha from 41
counties;
– Estimated budget: 159 million lei (VAT included);
– Period: 4 years -2015 -2018
– NACLR – National Organizer
– Participants: Experienced companies with sustainable
financial capacity.
Second level – Public auction for 1 lots with 1 -3
lots/counties:
– Estimated budget: 1,395 million lei/UAT (VAT
included);
– Period: 1 year;
– OCPI – Counties Organizer;
– Particip ants: Small -medium companies with
competent cadastral works skills.
Third level – financed by the local authority:
– Estimated budget: 60lei/parcels or 150 lei/hectare for
properties registered in e -CF;
– The cadastral works could be operated in small
regions – cadastral sectors;
– Small companies and all authorized person by
NACLR and OCPI.

At 31 January 2015, the current state regarding the
systematic registration of properties in Romania has reach
the following statistic situation:
– 7 cities of 320 has ongoing cadastral works funded
– from local budgets;
– 7 administrative territorial units has been completed,
with registration of 70.000 properties;
– 53 administrative territorial units has cadastral works
in progress, in different phases of implementation.

Journal of Geodesy, Cartography and Cadastre
20

Fig.10 – e-Cadastre and e -CF national program

6. Conclusions

Cadastral systems need to be flexible, because users
require daily updates on modifications of legislation,
technology and those regarding new means of
registration. Internet use, the use of GPS technol ogy,
exchange of data through XML and of modeling
standards of UML (Unified Modeling Language) data
have contributed over the past few years to the efficient
development of some viable cadastral systems.
Cadastral systems have developed over time all over the
world on the base of national or regional models, in
various forms, meeting different needs. [9] We
distinguish between: cadastral systems with fiscal
functions, accounting and territorial administration
systems, urban planning, environmental protectio n,
multi -purpose cadastral systems, land registration
systems. Nevertheless, the main functions of cadastre are
parcel representation and establishing the owner.
Presently, in Romania, cadastre is carried out
systematically and sporadically, while locally, at the
level of the territorial and administrative division, general
cadastre is still in the pilot project phase. In the future,
the plan is to connect the data bases belonging to
different institutions, namely the data base of the
Cadastre Agency to the data base of the National
Register of Personal Data.
One can thus find out in real time if a person is still alive,
if they possess the ownership right over a building
subject to cadastral operations.
Also for verification purposes we have the attempt to
connect to the data base of the Trade Register, thus
knowing whether companies are active, dissolved or in
insolvency.
Another important achievement would be putting together, at the level of each territorial and administrative
division, an electronic cad astral register of the street
nomenclature that could be accessed by both the agency
and the city halls.

References

[7] A.C. BADEA, Gh. BADEA, The concept of
cadastral system at international level , RevCad
Journal of Geodesy and Cadastre, Alba Iulia, 2005,
pp. 33 -43, GeoCAD `05 International Scientific
Symposium organized on 5th-6th of May 2005 by „1
Decembrie 1918” University.
[8] Gh. BADEA, A.C. BADEA, I. SION, Secțiunea
“Evidență cadastrală”, Sisteme informatice de
evidență cadastrală , vol. I, Curs postunivers itar de
perfecționare, Ed. Conspress, București, 2004.
[9] J. KAUFMANN, Cadastre 2014: From Theory to
Practice , FIG Working Week, Seoul, Korea, May 6 –
11 2001.
[10] J. KAUFMANN, Benchmarking Cadastral
Systems , FIG -Commission 7 – Cadastre and Land
Management – Reform ing the Cadastre, April 2002.
[11] D. STAUDLER, Modern Cadastre and Land
Administration , Tehran, 21 -26 July 2007.
[12] ***Permanent Committee on Cadastre in the
European Union – Cadastral Information System – a
resource of E.U. policies, part III , Romania,
Lantmater iet, edition 2009.
[13] A.C. BADEA, Gh. BADEA, C. DIDULESCU, G.
BǍDESCU, A. SAVU, Some Features of Project
Management Using Dedicated Software in the Land
Surveying Works , pp. 76 -81, MCBE'11 , Transilvania
University of Brasov, Romania, April 11 -13, 2011,
MCBE Session: Risk Management;
[14] G. BǍDESCU, O. ȘTEFAN, R. BǍDESCU, Gh.
BADEA, A.C. BADEA, C. DIDULESCU, Air-
borne Photogrammetric Systems Used in
Topographic and Cadastral Works in Romania ,
Recent Advances in Remote Sensing, Proceedings of
the REMOTE ’09 ,University of Genova, Italy,
October 17 -19, 2009, ISBN: 978 -960-474-129-8,
ISSN: 1790 -2769, Published by WSEAS Press
[15] C. GRECEA, G. RUSU, C.C MUȘAT, A.M
MOSCOVICI, Challenges in implementing the
systematic land registration in Romania , Recent
Advances in Geodesy and Geomatics Engineering;
Proceedings of the GENG '13 , pp. 98 -105; ISSN:
2227 -4359; ISBN: 978 -960-474-335-3, Published by
WSEAS Press;
[16] ***http://mycoordinates.org/wp –
content/uploads/2013/02/fig6 -12.jpg
[17] ***Government of Romania – Department for
Infrastructures Projects and Foreign Investment, The
Execution of the Cadastral Surveys and Real Estate
Publicity – Annual Report of IT Cadastral System in
Romania

Journal of Geodesy, Cartography and Cadastre
21
Deformation prediction and analysis by finite el ement method

Cătălin Ionuț Vintilă, Andreea Carmen Rădulescu, Petre Iuliu Dragomir (coordinator)
Rece ived: April 2015 / Accepted: September 2015 / Published : December 2015
© Revista de Geodezie, Cartografie și Cadastru / UGR

Abstrac t
This article presents the basic idea in the finite element
method, which is to find a solution of a complicated
problem by breakdown to several simpler problems. The
main issue is that since the actual problem is replaced by a
simpler one, it will be able to find an approximated solution
rather than the exact one, the approximation is strong
related to the type of the discretization and the number of
finite elements use to solve the problem.
Although this method has been extensively used in other
engineering fields like structural mechanics, fluid
dynamics, heat conduction and electrical field it can be
successfully applied to solve several problems in
deformation analysis.

Keywords
Finite element method, Prediction, Deformation Analysis

Ph. D. Stud. Eng. C. I. Vintilă
Faculty of Geodesy/ Technical University of Civil Engineering
Bucharest
Lacul Tei Boulevard, No. 124, District 2, Bucharest, RO
catalin.vintila@ymail.com
Ph. D. Stud. Eng. A. C. Rădulescu
Faculty of Geodesy/ Technical University of Civil Engineering
Bucharest
Lacul Tei Boulevard, No. 124, District 2, Bucharest, RO
andreeaa3radulescu@yahoo.com
Ph. D. Eng. Prof. P.I. Dragomir
Faculty of Geodesy/ Technical University of Civil Engineering
Bucharest
Lacul Tei Boulevard, No. 124, District 2, Bucharest, RO
petreiuliu.dragomir@gmail.com 1. Introduction

Deformation analysis involves studying the displacement of
network points in several measurement stages. Thus, there
are surveyed sizes as directions, distances, level
differences, the respective networks are compensated by
means of some methods in accordance with the
requirements, then the next step is the defor mation analysis
using the global congruence test and it is determined by the
application of some localization statistical tests. During
these tests, the limit values are determined with the help of
some tables depending on certain parameters and values for
each point. The two values are then compared and the result
obtained is used to determine whether a point is displaced
or not. In our days it surveying blocks can be analysed
performed in several stages, while at first just two models
were compared. [1].
If we have a geodetic network that was carried out in order
to obtain the coordinates of a set of geodetic points and
then used as a reference point for subsequent topographical
works, the measurements are carried out and processed only
once. In a geodetic network carried out to determine the
displacement or the deformation of lands or buildings, the
measurements are carried out and processed in several
stages between which there are determined the
displacements as differences in coordinates. In reference,
it's proposed the block processing of the measurements
carried out in several stages in a tracking network. [2]
The basic principle in the finite element method (MEF) is
finding a solution for a complex problem by replacing it
with a simpler one. By this r eplacement it can be found just
an approximate solution of the problem instead of the exact
solution. This method can be applied in the cases in which
no mathematical model can't solve the problem by finding a
precise solutions, but there are cases when th ere can't be
found nor approximate solutions by other methods. Thus, in
the absence of any other ways to bring about a resolution of
the problem, even if it's approximate, the finite element
method is a method which shall be taken into account. It

Journal of Geodesy, Cartography and Cadastre
22
has, on the other hand, the advantage that the solution can
be improved by applying a bigger volume of calculation by
dividing the problem in a larger number of finite elements
in order to obtain an approximate solution.

Fig. 1 The circle perimeter approximation using MEF

In the figure above the letterings represent:
– P, P(i), P(c) – the perimeter of the circle, the perimeter
of the polygon inscribed in the circle, respectively the
perimeter of the circumscribed polygon;
– r, s – length of side of the polygon inscr ibed,
respectively circumscribed in the circle.
When using the finite element method, the final solution for
the problem of the studied object is considered to be the
sum of the individual solutions of each element. A simple
example for the way in which th e finite element method can
be used is the approximation of the perimeter of a circle
with the perimeter of a polygon with n number of sides.
The polygon can be circumscribed as well as inscribed in
the circle, as it can be seen in fig. 1.
In order to demo nstrate that the perimeter of a circle is
approximated using the perimeter of a polygon with n
number of sides, it must be demonstrated that:
P)i(PnLim
(1)

P)c(PnLim
(2)

For this, the perimeter of the two p olygons must be
expressed in accordance with the radius of the circle, R, and
the number of sides, n. Thus, for each side of the polygons
it can be used the following relations:
n2Rsinr
(3)
n2Rtans (4)
Taking into consideration that the perimeter of a polygon is
the product between the length of side with the total
number of sides, it can be said that:
n2nRsin nr P(i)
(5)

n2sRtanns P(c)
(6)

The above relations can be transcribed in this way:
nnsin
R2 P(i)






(7)

nntan
R2 P(c)






(8)

As n tends to infinity, it follows that π/n tends to 0, so:
PR2 P(i) 
(9)
PR2 P(c) 
(10)

This demonstration shows that, despite the fact that the
finite element method is a method that has an approximate
result, the growth of the number of finite e lements is
directly proportional with the precision of the method. It
can be said that when the number of finite elements tends to
infinity, the method doesn't have an approximate solution,
but an exact one.

2. Historical Background

Even if the name of th e finite element method has been
recently assigned, the concept has dated for a few centuries.
For example, the calculation of the circle circumference
through approximation with the perimeter of a polygon was
realized by mathematicians of the Ancient time s. Thus, in
nowadays terms, each side of the polygon is a finite
element, and as the number of the finite elements increases,
the approximate value converges to the real value. These
characteristics are generally valid in every application of
the finite el ements. [6]
In order to find the differential equation of a minimum
surface area delimited by given closed curve, Schellback
discretized the area in many triangles and he used an
expression of a finite difference in order to calculate the
total discretized area in 1851. In the finite element method,
a differential equation is solved replacing it with a set of
algebraic equations.

Journal of Geodesy, Cartography and Cadastre
23
Since the 1900s, the behaviour of structural frames,
composed of many bars arranged in a regular way, was
approximated through th e behaviour of an elastic isotropic
body.
In 1943, Courant presented a method to determine the
rigidity at torsion for a tubular tree through the graduation
of the transversal section in many triangles and by using a
linear variation of the torsion functi on for each triangle in
its characteristic points (called node in the terminology of
this method). This study is considered to be the origin of
the finite elements method.
Since the 1950s, the aeronautics industry engineers have
begun to work at the devel opment of an approximate
method of prediction for the forces exerted by the wind on
aircraft wings.
The name finite elements was used, for the first time, by
Ray William Clough, considered to be the founder of the
method. Even if the method was initially d eveloped through
intuition and mathematical arguments, it was recognised to
be a form of the classical method Rayleigh -Ritz at the
beginning of the 1960s. As the same time as the recognition
of the mathematical base of the method, the development of
new me thods of the finite elements for different kind of
problems increased the popularity of the method.
The fact that both had a fulminant and IT development led
to the validation of the method in a practical and can be
achieved high volume calculation method that is part of a
more or less automatically.
A broader interpretation of the finite element method was
carried out by Zienkiewicz and Cheung and applicability of
the method in any field. With this interpretation has been
calculated that extended finite e lement method can be
derived by applying a weighted residual methods such as
Galerkin method or the method of minimum squares. This
expanded interest mathematicians and engineers in
applying the finite element method for finding the solution
of various lin ear or non -linear differential equations.
Traditionally, mathematicians are those who develop
techniques and methods for finding solutions of differential
equations and engineers use these methods to solve
problems in various fields. Only if the finite ele ment
method are the engineers who developed the technique for
mathematicians to use the method for finding the solution
of complex differential equations. Nowadays, solve
engineering problems by finite element method has become
an industry, using the metho d allowed millions of degrees
of freedom.
The rapid progress of the finite element method can be seen
by the fact that annually about 3,800 research papers are
presented and published. Since 1995 were estimated a total
of 56,000 publications, 380 books and 400 conferences in
the finite element method.
By this progress today finite element method is considered
as the most well -established and convenient method of
analysis by both engineers and the scientists.
3. Presentation of the method

The finite element method is a method of processing
approximate differential equations partial differential
equations that describe physical phenomena or not real.
In principle, finite element analysis is to divide the area into
zones of simple geometric form, their analysis and
recomposition domain under certain mathematical goals.
Areas of application of the finite element method can be
any fields that describe a phenomenon using differential
equations. Until now, the method has developed
remarkably in areas such as fluid a nalysis, analysis of
electrical, magnetic analysis, thermal analysis, and
structural analysis. [6]
Calculation program used to analyse the problem does not
solve the actual structure, but its models. Modeling is an
activity to simplify the structure by try ing to fit various
portions of simplifying structures and bearing loads.
Modeling is directly related to the model and is very much
everyone's experience, inspiration and knowledge is closely
related to the theoretical basis of the method. Every
program th at is based on the finite element method has
some peculiarities that must be learned but there are some
basic rules that once mastered the method allows
addressing any program that is based on the finite element
method.

Fig. 2 The processes performed be fore execution

Journal of Geodesy, Cartography and Cadastre
24
The finite element method has developed into an
indispensable technology in modeling and simulation of
advanced engineering systems. In an attempt to build an
advanced engineering, engineers and designers go through
a sophisticated process w hich consists of modeling,
simulation, visualization, analysis, design, testing and
finally production. Most often a huge amount of work is
done before the actual manufacture of construction. This is
done to ensure the optimum use of the construction, but
also to have a low cost of the construction project. This
process is illustrated in Figure 2, the process that is taken
up several times in different phases due to the data obtained
in the field. The techniques that are directly related to
modeling and sim ulation plays an important role, and finite
element method is a standard tool used in almost any large
scale project.
The finite element method was initially used to solve
problems of structural analysis and analysis of solids and
has since been used to so lve different types of problems in
various branches of engineering. In principle, the finite
element method is a numerical method that seeks an
approximate solution distribution in the investigated
variables, which is usually difficult to obtain analytical
methods. Apply by dividing the domain into a finite
number of pieces called elements usually simple shapes in
geometrical point of view.
Figure 3 illustrates schematically the distribution of a
function f (x), where it is approximated using
unidimensional finite element method. In this case f (x) is a
continuous function is approximated using piecewise linear
functions in each item. If unidimensional end of each
element is called node. Unknown variables in finite element
discrete function values are in kno ts. Mathematical
Principles are so used to establish equations for each
element, then the elements are related to each other to
describe the distribution of the whole geometry. [6]

Fig. 3 The finite element method approximation
Steps to solve a problem using the finite element method.
1. Dividing the domain of finite element analysis
At this stage, choose the type of finite elements according
to the subject studied and divide the structure into finite
elements. This operation is called meshing and often is
done automatically by computer. Finite element type is
defined by several aspects such as the number of
dimensions (one -dimensional, two -dimensional or three –
dimensional); number of nodes of each element;
Approximation functions to be associated. The q uality of
results is directly influenced by the choice of finite
elements and the way it is meshed object studied .
2. Establishment of finite element equations
The behaviour of the studied object in the body of a finite
element equations described by finit e elements, also called
elemental equations. They form a system of equations for
each finite element.
3. Assembling the elemental equations in the system of
equations of structure
The behaviour of the whole structure is modelled by
assembling systems of eq uations for each finite element in
the overall system of equations of the structure, which
means that the balance is directly proportional to the
equilibrium structure of finite elements. By joining their
common nodes requires that the unknown functions ha ve
the same value.
4. Implementation of boundary conditions and solving the
structure system of equations
The system of equations thus formed is solved to obtain the
values of the functions in each node of the network. These
are called primary or first -order unknown.
5. Perform additional calculations to determine the
secondary unknown
In some cases, after determining the primary unknowns, the
analysis ends. This is usually the case for problems where
the unknowns are to be found in the nodes formed by
mesh ing. There are cases, however, the primary unknowns
just knowing is not enough, the analysis continues to
finding some unknowns secondary or second order. Thus,
for example, in monitoring deformations of primary
unknowns are the displacements of nodes. Wit h their side is
determined the secondary unknowns that are specific
deformations in every node.

4. Mathematical model

A mathematical model for the analysis of a structure
involves automatically determining a number of variables
and functions u*(ξ) representing the displacements, strains,
ξ coordinate the functions ξ (x, y, z), defined nodal points. If
the exact solution u*(ξ) is unknown and u (ξ) is an
approximation of it, where the error function e (ξ) is:
)(*u-)u()e( 
(11)

Journal of Geodesy, Cartography and Cadastre
25
To find an approximate solution of the problem is
sufficiently to describe an expression which contain n
parameters of approximation.
)a,…a,a,u()u(n 2 1
(12)

This will determine these parameters based on the error
function.
The finite eleme nt approximation is nodal and has the
following form:
 nT
n321
N 3 2 1 uN
uuuu
)(N… )(N)(N )(N)u( 






(13)
Where:
– ui are nodal parameters of approximation and has a
concrete physical significations (nodal displacement),
– Ni(ξ) – interpolation or approximation functions, which
usually has polynomial forms.
The approximate solutions u(ξ) can be built on the entire
structure domain of definition V or on elementary
subdomains Ve of the finite elements which means that:


n
1eeV V (14)

Each finite element is numbered, usually from 1 to the total
number of elements n, reference to an element is the index
e., The operation of a field in a finite number of nodes and
finite elements, whether the finite eleme nts are alike or not,
is called meshing.
Meshing domain analysis is the first step to be performed in
numerical solving a problem using the finite element
method. Meshing involves deciding principle forms the
finite element size and type, but all these are strictly related
to the element studied.
Each type of finite element should be designed to satisfy
the most closely studied the mesh object and there is not a
finite element "generally valid", with which you can model
any geometric shape. [3]

5. Meshing do main analysis

Engineers are often faced with the delicate problem of the
relationship between the numerical solution quality and
computational effort made to obtain the solution. In
general, the mesh size is directly proportional to the
accuracy of the so lution. However a large network of finite
elements leads to large systems of equations, whose
solution is often difficult. This is particularly important in
determining the number of finite elements mesh mode
domain analysis the analyst has to be, by exper ience, to
decide on increasing the mesh only in areas where large
gradients are expected of unknowns. In areas where such
variation is considered small to insignificant impact can be achieved minimal networks of finite elements so that the
result is about the same accuracy. [6]

Fig. 4 Analysis domain discretization

Each node of the mesh thus formed shows a possible
displacement on ox axis and on oy axis, so one can say that
there are two parameters that define a uniquely moving a
node in two -dimensional space. These parameters are called
degrees of freedom of a node, in literature degrees of
freedom are denoted by DOF .
The calculation model of a structure to be subjected to an
analysis by finite element method, where generally consists
of lines, flat or curved or volumes. At this stage of
modeling, the model is continuous with an infinite number
of points. Meshing is the fundamental approach in the finite
element method and consists of converting a continuous
structure with an infinite number of points, a discrete
model, model characterized by a finite number of points,
called nodes.
The nodes are defined primary nodal unknowns whose
values are calculated by finite element method.
Uncertainties associated nodes can be trips, where the finite
element model is called displacement. To shift the pattern is
recognized that the deformed shape of a structure, as a
result of the various requirements, is defined by the
displacements of all nodes in relation to the network node
prior to deformation. Each node can hav e up to six
components of the movement, this being called nodal
displacements, with respect to a reference coordinate
system and the system to which it is reported the overall
structure of the three components x, y, z, and three rotations
of the linear mov ement Φx, Φy, Φz.

Journal of Geodesy, Cartography and Cadastre
26
6. Types of finite elements

A general classification , rather vague classification, finite
elements divided into three categories , depending on the
number of dimensions which constitute the mesh in which
it is used :
– Unidimensional finit e elements;
– Two-dimensional finite elements;
– Three -dimensional finite elements.
Unidimensional finite elements are used when the size
considered to be approximated depend on a single variable,
so that the items are one -dimensional line segments or arcs
curve along which takes values independent variable of the
problem.

Fig. 5 Unidimensional finite elements

Dimensional finite elements are used when the approximate
size directly depends on two variables. For example, they
can be used to resolve the two -dimensional elasticity,
deformation and plane displacement.

Fig. 6 Two -dimensional finite elements

Three -dimensional finite elements form a fairly large class,
from this class increasingly common are the following two
families of finite elements: tetrahed ral finite elements and
hexahedral finite elements.

Fig. 7 Three -dimensional finite elements

In practice are rarely used elements function approximation
of degree greater than three. The great diversity of elements
currently in use leads to inability to find a single criterion
for classification. In particular, high mathematical
complexity finite elements are the most difficult to include
in a certain class of items.

7. Advantages and disadvantages

The main advantages of the finite element method are:
– Flexibility – this feature is given by the fact that it allows
meshing of bodies or surfaces no matter how complex these
are in elements of geometric shapes not only simple, but
also very popular, and by applying basic math calculations
can calculate local variables for each finite element
separately which will be use them in determining global
variables and thus solve the problem and solution.
– The possibility of modelling the inhomogeneous bodies
considering their physical properties
– Ease to implement finit e element method in general
computer programs.
– The possibility of applying the same method for different
types of problems in various engineering fields due to the
generalization of the method .
– The simplicity of the basic concepts of the finite element
method can be considered an important advantage of the
method. The importance of ownership and understanding
them correctly may result from the fact that such concepts
include certain assumptions , simplifications and
generalizations . But ignoring these concep ts can lead to
serious errors in modeling and finite element analysis and
thus can become a disadvantage of the method.
The finite element method disadvantages:
– A major disadvantage of the finite element method can
be considered the large number of input d ata, the large
amount of computation to be performed to calculate the

Journal of Geodesy, Cartography and Cadastre
27
model. The largest share of the workload is meshing the
object into finite elements and calculating the coordinates
of nodes and the configuration of finite elements.
Currently processin g and analysis programs by finite
elements method on the market have specialized modules
for automatic discretization, thus relieving the user from
making a large amount of work by manual meshing.

8. Conclusions

The finite element method is a very popular analysis method
and it is used in several engineering domains. This method
presents the advantages that the studied object can be
discretized in simple finite elements, no matter how
complex the object is, the possibility to apply this method in
several d omains and the simplicity of the base concepts of
the method. On the other hand, the breakdown of the object
in an extremely large number of finite elements, to get
closer to the true result, lead to the formation of an equally
large number of nodes and he nce an extraordinary volume
of calculation.
If the computer technology didn’t evolve at the same time
with the method, facilitating the huge work, the lack of
precision of determination would made the method useless
for many branches of engineering, includ ing engineering
geodesy measurements.
The objects discretization in finite elements is the part where
the largest volume of work is filled, this part is easier with
specialized programs that have implemented software which automatically performs mesh, rega rdless of the complexity
of the studied objects.
The advantages of the finite element method, described
above , provides the opportunity to study this method not
only in predicting deformations or displacements but also in
processing and analyzing measurem ents.

References

[1] Vintilă, C.I., Rădulescu C.A., Dragomir P.I. –
Monitoring Vertical Displacements by Means of
Geometric Levelling for Sky Tower Building from
Floreasca City Center, Bucharest, GENG14, pag 31 -40,
Brașov, România, 2014
[2] Vintilă, C.I., Rădulescu C.A., Dragomir P.I. – Modern
techniques of lands and constructions deformation
monitoring, GeoCAD, Alba Iulia, 2014
[3] Ghergheles, L., – Analiza proceselor dinamice ale
constructiilor si terenurilor, Teza de doctorat, Bucuresti,
2011
[4] Rao, S.S. – The finite element method in engineering,
Editura Butterworth – Heinermann, Burlington, USA,
2011, ISBN 978 -1-85617 -661-3
[5] Liu, G.R., Quek, S.S., – The finite element method, a
practical course, Editura Elsevier, Croydon, UK, 2013,
ISBN 978 -0-08-098356 -1
Comșa, D.S. – Metoda Elementelor Finite, Curs
Introductiv, Editura U.T. P

Journal of Geodesy, Cartography and Cadastre
28

Testing a new method of digital images acquisition in the process of
3D reconstruction of an object

Paun Corina Daniela, Valeria Ersilia Oniga, Ciobanu La ura Elena
Rece ived: April 2015 / Accepted: September 2015 / Published : December 2015
© Revista de Geodezie, Cartografie și Cadastru/ UGR

Abstrac t
In more and more spheres of activity (industry,
constructio n, aerospace, quarry, etc.) it is necessary to
achieve three -dimensional models. Currently, using high –
resolution digital images for the 3D reconstruction of an
object can be a precise and an inexpensive alternative,
used more frequently in the research st udies. The data
acquisition process is performed without any direct
contact with the studied object, in a very short time and
the results of processing the acquired digital images are
reliable and accurate. In most of the software, for the best
results, it is recommended to take the images so that the
angles between the skew rays to be as close to right
angles as possible, while moving around the object to be
reconstructed in 3D. In order to test the functionality of
this new method of images acquisition, t here was created
the 3D model of a stone, using the images acquired with
the Canon EOS 60D digital camera. The stone’s 3D
model was represented as a point cloud which was
automatically generated using the Structure from Motion
(SfM) algorithm. Then, the po int cloud was scaled based
on a reference model obtained by TLS measurements, i.e.
a reference distance. Finally, the point cloud obtained by
close -range photogrammetry was compared, using the
Hausdorff distance, with the point cloud obtained by TLS
techno logy. The standard deviation of the distances
between these two point clouds is a measure of the
accuracy of the point cloud obtained by using the
proposed configuration of digital camera positions.

Keywords
digital image, image configuration, SfM, preci sion

PhD. Candidate Ing. Corina Daniela Paun
Faculty of Geodesy
Technical University of Civil Engineering Bucharest
Address: Lacul Tei Bvd., no. 122 -124, sector 2, Bucharest
E-mail: dcpaun @yahoo.fr 1. Introduction

Currently there is an interest in me asuring stone blocks
derived directly from quarries or mines, because some
applications require stone blocks with default sizes and
shapes. It is also necessary that each block stone
extracted from the quarry to be used to its full potential
without any re sidual materials.
At this time, in order to fulfil the requirements about the
blocks stone, the performed measurements are kept
simple. This involves additional workforce and the
results are minimal and can't be used in most of the
applications.
The aim of this paper is the implementation of a new
method for the 3D reconstruction of a block stone
coming from a quarry, using close range
photogrammetry, by combining two images
configuration (viewing directions for object
measurement) for the same object: para llel and
convergent. Based on this method, the dimensions of the
block stone will be obtained in a completely automatic
way, being able to estimate its volume.
With the development of computer visualization, the
current interest is to get all the informati on in digital
format, with a minimum human interaction and high
precision, so that their processing in order to obtain
additional information to be done in fully automatic
mode.
Traditionally, photogrammetry has been defined as the
process of deriving metr ic information about an object
through measurements made on photographs of the
object [Mikhail M. et al. ,2001]. Usually, it involves the
analysis of one or more existing photos or video
sequences, using specialized photogrammetric programs
to establish spa tial relationships. The measurement
method is without any direct contact with the studied

Journal of Geodesy, Cartography and Cadastre
29
object, the results are accurate and reliable, data
collection is made in a short time and with low cost, the
images are taken and stored, can be consulted and
remeas ured anytime in the future. This technology is in
the pipeline and presents an alternative to the
measurements made by the terrestrial laser scanner.
The literature in the field includes many algorithms for
the digital images analysis and processing, and t he best
known and used is the algorithm Structure from Motion
(the process is fully automated), which is based on
bundle adjustment and image matching algorithms.

2. Image acquisition

The image acquisition process may depend on the
available equipment an d the required results (high or low
resolution and accuracy) by applying different
photogrammetric techniques.
In photogrammetry, imaging configuration is referred to
as the arrangement of the camera stations and viewing
directions for object measurement. Usually in
photogrammetry, the following imaging configurations
are distinguished: single image acquisition, stereo image
acquisition and multi -image acquisition [Luhman n T. et
al., 2006]. There are three cases for stereo image
configuration: parallel (or normal case), convergent and
shifted (different scale for both images). It is generally
important to achieve approximately 60% overlap
between images with a minimum of 4 control points in
the overlap area.

(a) (b)

(c)
Fig. 1 Stereo image configurations (a) parallel, (b) convergent and
(c) shifted [modified after Linder, 2006]

The disadvantage of the convergent case is that
additional perspective distortion is produced. Normally,
the parallel case is suitable for stereo vi ewing and
automatic surface reconstruction. The convergent case
can often lead to a higher precision particularly in z
direction. For height measurement, the photos are
preferably taken with a short focal length (zoom off) camera in order to get a large B/ h -ratio [ Singh V. P. et
al., 2011].
The traditional image acquisition by rotating the camera
around the object isn’t suitable for this case study ,
because an operator should be involved in taking the
images all along and the acquisition time would be ver y
long, while the main objective is the automation of the
process.
In this stage of the study, the working team thought about
an efficient method for the digital images acquisition,
which would be reliable under the conditions of quarry,
but also to provid e precise results. Thus, it was
determined that the best way for data acquisition
supposes that the stone blocks are transported on a digger
platform which will pass between two benches that will
have two cameras installed at the upper ends. These
cameras will take images of the top of the block stone,
side parts and front part (Fig. 2) .

Fig. 2 Digital image acquisition mode in the quarry
For this study we used a block stone from a granite
quarry, with the dimensions of 100 x 110 x 54 cm and
the simulatio n of the process which will take place in the
quarry, was made in the basement of the Faculty of
Mathematics and Geoinformation, Technical University
of Vienna, during a training stage through the Erasmus
program. For this purpose, the stone was encased on a
pedestal with wheels, to simulate the displacement. Two
tripods were used to foster the reduction of the blurr ied
images, on which there were installed two digital
cameras, Canon EOS 60D. The position of this two
tripods was chosen so that the distance, towards the axis
on which the stone will be moved, to be almost equal.
During the image acquisition process, the stone was
moved on a distance of approximately 7 m, in straight
line (Fig. 3).

Fig. 3 The image acquisition process simulation in the ba sement
of the Faculty of Mathematics and Geoinformation

Journal of Geodesy, Cartography and Cadastre
30
The two Canon EOS 60D digital cameras parameters, are
shown in Table 1:
Table 1 The Canon EOS 60D parameters used for image acquisition

Resolution 18 MP (5184×3456 pixels)
Focal length 20 mm
Prog ram Manual
Focus mode Manual
Digital zoom None
Aperture F 8,00
ISO 1000

The concept used to capture images is called multiple –
image -photogrammetry [Singh V. P. et al., 2011] and it
combines the converged position case of the cameras (for
matching poi nt pairs in the front and top of block stone)
and the parallel case (for point pairs in the sides of the
block stone). There were taken images with each camera
in such a way so that there should be an overlap between
both images acquired with the same came ra and also
between images from different cameras. Through this
process there will be made the connection between
images and the relative orientation of the cameras in
relation to the photographed object (Fig. 4).

(a) (b)
Fig. 4 (a) Image configur ation for the block stone 3D reconstruction,
(b) example of two images taken with the two cameras used to
calculate the coordinates of the upper right corner of the block stone

3. Tridimensional reconstruction process of
the block stone

The traditional photogrammetric methods suppose
knowing the spatial locations and orientation angles of
the camera used for data acquisition and the existence of
control points whose 3D coordinates are known, in order
to reconstruct a scene in 3D space . In contrast, the
methods based on SfM resolve automatically and
simultaneously the problem of camera position and the
geometry of the scene, by using some common
characteristic points of the successive images which
present a considerable overlap.
A pair of images with a min imum of 60% overlap
containing the same object, taken from two different
viewpoints, can be used to create the stereoscopic model
of the object . So, b y using stereoscopic display techniques , we are able to view the model
stereoscopically, to measure the th ree-dimensional object
coordinates and to digitize objects like points, lines or
areas, sometimes called feature collection, in order to
reconstruct the object in 3D [Linder W., 2006] .
Instead of a single pair of stereoscopic images, SfM
involves the use o f several images with sufficient
overlap , from which there will be extracted the feature
points that will be used for the 3D reconstruction of the
scene.
The block stone 3D reconstruction process includes the
following steps [modif ied after Ruther H. et.al, 2012 ]:
– filtering the images by removing the background;
– calculating the spatial locations and orientation angles
of each camera (exterior orientation) using the bundle
adjustment process;
– computing the image matching process;
– calculating the co ordinates of the point cloud (local
projection) extracted from images by bundle adjustment
process;
– analyzing the accuracy of the point cloud obtained after
image processing by comparing it with the TLS point
cloud using the Hausdorff distance.

3.1. Ima ge filtering

While taking the photos, the object that is intended to be
reconstructed can't cover 100% of the image area, for
which reason, the undesirable objects in the background
will appear in the images, respectively in the point cloud.
This results in a high workload for point cloud modeling,
because the filtering process can't be fully automated. For
this reason we resorted to the stage of removing the
background from the images and this stage implies both
automatic and manual methods.
For this case study, removing the background from the
images was achieved in fully automatic mode. As
reference images there were used two images taken by
two cameras, which show only the background from the
environment in which the images were taken and used
for the 3 D reconstruction of the stone (Fig. 5) . These
images show the static objects that will not change
during the takeover of the entire series of photos.

(a) (b)
Fig. 5 The reference images for the background removal

In order to obtain the results, a script in the MATLAB
programming language was written. Firstly, a directory

Journal of Geodesy, Cartography and Cadastre
31
was created, containing the images from which the
background needs to be removed.
The order of operations performed is as follows:
a) Importing the images containing only the ba ckground;
b) Importing the images containing the background and
the object that is intended to be reconstructed, in this case
the block stone (Fig. 6a);
c) Conversion of the images that contain information
about the RGB code into images with gray values. A t this
stage the hue and saturation information are eliminated,
while retaining the luminance;
d) Calculation of differences between the images
including only the background and those that contain the
block stone as well (Fig. 6b). Because the stone is
moving, there were some errors in this stage,
representing pixels or groups of pixels belonging to the
background and for which the difference values are not
zero or close to zero;
e) Creating a mask for transforming the images which
contain the absolute dif ferences in binary images. To
determine more precisely the area that is intended to be
preserved from the images which contain the block stone,
it was used a threshold of 50 for the absolute difference
values. The resulting images will contain white pixels
corresponding to the absolute difference of more than 50
(these will be retained), and the black pixels
corresponding to the absolute differences of less than 50,
representing the areas to be removed (Fig. 6c);
f) Removing small groups of pixels from bina ry images.
In this way, the accuracy of selecting the pixels that will
be kept in the final images is improved. After several
tests, it was established that all the objects connected
with more than 1000 pixels will be removed (Fig. 6d);
g) Selecting the re gions that contain only pixels that will
be retained. Thus, cutting the images keeping only the
regions that contain the wanted pixels (Fig. 6e);
h) Transforming the binary images into some images that
contain information about the RGB code. At this stage, it
returns to the color images, only for the regions that have
been selected in the earlier stages (Fig. 6f);
It is noted that there can't be achieved a complete
elimination of the background from the images, because
there are elements that are not part o f the background,
but they change their position while moving the block
stone. However, it was removed a high workload and the
filtering of the point cloud after removing the
background can be achieved in a much shorter time.

(a) (b)

(c) (d)

(e) (f)
Fig. 6 The steps used for the background removal

In the future, for the full automation of the process for
removing unnecessary points, it is proposed an algorithm
based on the automatic identification of unwanted pixels
following the RGB c olor codes. Since the stone texture
provides close color pixels, the background will be easily
distinguished from the block stone. It is observed that in
the cropped images there can be found the calibration
grid (checkerboard). If it is desired to do the calibration
process on the job, this will be stored and used to
determine the parameters used for the interior and
exterior orientation of the cameras. If the calibration will
be done separately, the checkerboard will be at its turn,
detected and removed from the image. Also, based on the
checkerboard of known size, the scaling of the block
stone can be achieved, as on digital images from non –
metric cameras, the reconstruction of objects at real size
can't be done.

3.2. Structure from motion algorithm

For the 3D reconstruction of the block stone there were
used the open source Visual SfM software and 23 images
from which the background was removed.
The development of this technique began in 1990
[Spetsakis ME., Aloimonos Y. , 1991] in the field of
computer vision, while the study of algorithms for
automatically finding the characteristic points began a
decade before.
Using the structure from motion algorithm, t he 3D point
cloud is generated in a local coordinate system "image –
space", which must be aligned w ith a real coordinate
system "object -space". In most cases, this transformation
can be achieved using a small number of ground control
points having known spatial coordinates. These points
can be identified both in the point cloud and on terrain,
and their coordinates can be determined by terrestrial
measurements. However, in practice, it is often easier to
install on the ground, before the image acquisition,
targets with high contrast and the center well defined.
This approach simplifies the correlation o f image space

Journal of Geodesy, Cartography and Cadastre
32
with object space and also provides a high degree of
confidence. A good distribution of network targets on the
study surface offers an accurate assessment of errors
produced by reconstruction based on SfM. Visual SfM is
an GUI (Graphical User Interface) application for 3D
reconstruction which is based on structure from motion.
The geometrical theory of structure from motion allows
projection matrices and 3D points to be computed
simultaneously using only corresponding points in each
view [ Robe rtson D.P. and Cipolla R, 2009 ]. Structure
from motion techniques are used in a wide range of
applications including photogrammetric survey [Kraus
K., 1997], the automatic reconstruction of virtual reality
models from video sequences [Zisserman A., et al.,
1999], and for the determination of camera motion (e.g.
so that computer -generated objects can be inserted into
video footage of real -world scenes) [Kutulakos K. N.
et.al.,1998].
3.2.1. Image matching
The first step is to find common characteristic points and
the second is to correlate the images based on them.
Using the images where the background was removed,
the software found 37.405 feature points, based on which
it made the correlation between 253 pairs from images
within 16 seconds. The software show s the number of
characteristic points for each image and the time needed
to determine them.
During the process of image matching it was observed
that the highest number of points is identified in the
second image, respectively 3197 feature points, because
in this image the stone is very close to the camera and
occupies a considerable area of the image (Fig. 7a). For
the last image, we see that the stone occupies only a
small part of the photo (Fig. 7b).

(a) (b)
Fig. 7 (a) Second image from the 23 set of images and (b) the last
image

This operation was also done for all the images that
contain the background, and the program has recognized
53 725 feature points in 29 seconds. Although the
difference in time isn't so large, this changes considerably
with concomitant use of multiple images or higher
resolution.
For each pair of images it can be viewed how the program
makes the correlation between them, looking if errors
occur during this process. In Fig. 8, we can see the pairs
of correlated points bet ween two images representing the front of the stone, the correspondence between images
defining the image matching process.
In other words , digital image correlation determines the
conformity between image parts of two or more images
that capture the same object at least in part, on the basis of
statistical analyses. Image correlation is a subs titute for
natural stereoscopic vision. It is based on the calculation
of correlation function in window with searched image
and pattern image [ Krajňák M. et.al. , 2011].

Fig. 8 Correlation of pair points

3.2.2. Bundle adjustment
In many cases, using a single pair of stereoscopic images
is not enough for the reconstruction of complex scenes.
Therefore, a large number of images must be used in
order to capture the object or scene from every angle.
Given a set of images for which the characteristic points
were determined, the role of the bundle adjustment
process is to find their 3D position and camera
parameters that minimize errors redesign points.
Almost everything that we will say can be applied to
many similar estimation problems in vision,
photogrammetry, industrial metrology, surveying and
geodesy [ Triggs B. et al., 2000 ].
This method is widely used because it offers many
advantages such as: flexibility (it processes a very wide
variety of different 3D features and camera types),
accuracy (it gives precise and easily interpreted results)
and efficiency (it uses economical and rapidly
convergent numerical methods and make near -optimal
use of problem sparseness) [ Triggs B. et al., 2000 ].

Fig. 9 The principle of bundle adjustment [Luhmann T. et al., 2006]

Journal of Geodesy, Cartography and Cadastre
33
The name of this method relates to the packages of rays
of light that come from each 3D characteristic point and
converge in the o ptical center of each camera, which are
optimally adjusted respecting the structure and also the
view parameters. This name originated from the 1950’s
and has been increasingly used in computer vision
research in recent years.
The mathematical formulation of bundle adjustment
(Fig. 9) is based on the co -linearity condition
[McGlone, 1989]:



11 12 13
31 32 33
21 22 23
31 32 33j ci j ci j ci r r r
ji pk i
j ci j ci j ci r r r
j ci j ci j ci r r r
ji pk i
j ci j ci j ci r r rX X Y Y Z Zr r rx x x fX X Y Y Z Zr r r
X X Y Y Z Zr r ry y y fX X Y Y Z Zr r r      
      
where:
xji, yji – image coordinated of point Pj on image i;
fi – principal distance of camera k at station i;
xji, yji – principal point coord inates of camera k;
Δx, Δy – correction terms allowing for systematic
errors;
r11i,.., r33i – elements of rotation matrix of image i;
Xj, Yj, Zj – object point coordinates;
Xci, Yci, Zci – perspective centre coordinates.

For this case study, following the bundle adjustment
process, the " VisualSfM " software, calculated the three –
dimensional coordinates of the characteristic points
identified by image matching algorithm , in a local
coordinate system , the interior orientation parameters
and the exterior or ientation for each camera position.
So, i n Fig. 10a we can see the point cloud representing
the block stone obtained by processing the original
images, containing 53725 points and in Fig. 10b, we can
see the point cloud representing the block stone
obtaine d by processing the filtered images (without
many part s of background ), containing 37405 points .

(a) Point cloud representing the block stone obtained by processing
the original images

(b) Point cloud representing the block stone obtained by proc essing
the filtered images

Fig. 10 Comparison between the point clouds representing the
block stone obtained before and after the background subtraction

The high difference between the number of points is due
to the fact that when using original image s, many point s
that don't belong to the block stone were calculated in the
bundle adjustment process. Thus, in the case of using the
images without background, the process of point cloud
filtering require s a shorte r time in comparison with the
case of usin g the original images .
In Fig. 11 we can see the relative orientation of the
images acquired using both cameras in relation with the
automatic ally generated point cloud. This process was
realized using VisualSfM software.

Fig. 11 Relative orientation of the 23 digital images in relation with
the point cloud

3.3. Scaling the point cloud representing the
block stone

Because on the block stone there were not any control
points with known coordinates (targets), which could
help to achieve the scale of the stone, we used the point
cloud obtained by TLS measurements in order to achieve
this stage . So, the point cloud was scaled based on the
TLS point cloud, i.e. a reference distance (Fig. 12), using
the “Geomagic” software. The scale factor was
calculated as a ratio between the distance measured
between two TLS points and the distance measured
between the same points in the image point cloud.

Journal of Geodesy, Cartography and Cadastre
34

Fig. 12 Scale factor calculation based on the two point clouds (TLS
point cloud and automatically generated poi nt cloud based on digital
images)

Then, the two poin t clouds were brought in coincidence
based on three common characteristic points, using the
„Geomagic” software , in order to compare them.

4. Precision evaluation of the point cloud
generated using the SfM algorithm

To determine the accuracy of the point cloud
automatically generated by SfM algorithm, this must be
compared with a reference model considered for this case
study, the TLS point cloud .
Based on the TLS point cloud, a mesh surface was
create d, using the Poisson Screened su rface
reconstruction method, the co mparison being made
between the point cloud automatically generated by SfM
algorithm and the mesh surface, using as comparison
metric the Hausdorff distance implemented into the
“CloudCompa re” software.
The Poisson Screened surface reconstruction method was
chosen to create the mesh surface, because it fits best the
point cloud, a standard deviation of +0.076 cm being
obtained after comparing it with the TLS point cloud.
Hausdorff Distance – named after Felix Hausdorff, is the
most famous metric for comparing two mesh surfaces,
providing a global comparison [ Oniga V. E., Chirilă C.,
2013 ].
Considering two surfaces S and S’, the distance between
S and S’ is dS(S,S’) , defined as:
SpSd ( S,S') maxd( p,S'),

(1)
where:
p' S'd( p,S') mind( p, p')

and d(p,p’) is the Euclidean
distance between two points in R3.
This distance is not symmetric, for e xample d(S,S’)≠
d(S’,S) . The d(S,S’) distance will be called the forward
distance , and the d(S’,S) the backward distance . The
term Hausdorff symmetric distance ds(S,S’) , will be
introduced, defined as follow [ Aspert N . et al., 2002 ]:
ds(S,S’)=max[d(S,S’), d(S’,S)] (2)
The symmetrical distance offers a more accurate
measurement of the differences betw een two surfaces,
because the one -side distance can lead to an
underestimation of the distance values between the two
surfaces [ Aspert N. et al., 2002 ].
When the orthogonal projection p’ of the point p on the
triangle surface T’, is inside the triangle, th e distance
from the point to the triangle is nothing but a point to
plan distance. When the point projection is outside the
triangle T’, the point to triangle distance, is the distance
from the point p to the closest point p” that belongs to a
triangle sid e, as you can see in Fig. 13.

(a) (b)

(c)

Fig. 13 (a) The distances between S and S’ surfaces, (b) The
distance between p and T’ with p’ situated outside the triangle
[Aspert N. et al., 2002], (c) The distance between p and T’ with
p’ situate d inside the triangle [Roy M. et al., 2004]
Even if the distance d(p,S’) can be calculated analytically
for any point p, a sampling of the triangle surface is
necessary, in order to obtain the maximum for p S.
Thus, each triangle which belongs to S surface is
sampled, computing the distances between each nod of
each sample and S’ surface.
The distance accompanied by its sign was introduced in
the “ CloudCompare ” software for an independent
evaluation of the areas that belong to the first surface and
are situated inside or outside the space, relative to the
second surface [ Cignoni P. et.al., 1998 ].
After comparing the point cloud automatically generated
by SfM algoritm , with the mesh surface created based on
the reference point cloud (TLS point cloud) , the
following differences were obtained (Table 2).

Table 2 The numerical differences between the point cloud
automatically generated based on digital images and the reference
model, namely the mesh surface created based on TLS point cloud

Maximum positive : +5.16 cm
Maximum negative: -2.51 cm
Mean: +0.43 mm
Standard deviation +0.74 mm

Journal of Geodesy, Cartography and Cadastre
35
The differences are also highlighted using a color palette ,
as you can see in Fig. 14.

(a)

(b)

Fig. 14 (a) The differences between the point cloud automaticall y
generated based on digital images and the reference model, namely
the mesh surface created based on the TLS point cloud, (b) the
differences distribution histogram

Analyzing the color palette we can easily obtain the
value for the Hausdorff distance . So, the points situated
on the top and front parts of the block stone 3D model
are colored in green tone (Fig. 15a) , the differences being
between -0.5 cm and +1.3 cm and in blue tone, the
differences being between -2.5 cm and -0.5 cm. On the
other hand, the points situated on the sides of the block
stone 3D model are colored in green, yellow and orange,
the differences being between -0.5 cm and +2.5 cm (Fig.
15b).
Analyzing the histogram, we can say that for most of the
points the differences are in range of -0.1 cm and 1.8 cm ,
namely for 24122 points, and for a small number of
points the differences are situated outside the range,
namely for 1178 points, representing 4.6% of the total
number of points.

(a)

(b)
Fig. 15 The Hausdorff distances calculate d for two points which
belong to the image point cloud, the points being situated (a) on the
block stone top part, (b) on the side of the block stone

5. Conclusions

The accuracy of object 3D reconstruction process is
directly proportional to the accuracy of determining the
scale factor and t he accuracy of the precision evaluation
process also depends on bringing in coincidence the
model to be evaluated with the reference 3D model.
When creating a 3D model as a point cloud by using the
Structure from Motio n algorithm, the result is obtained
with high accuracy, in a completely automated way and
in a very short time. By comparing it with the reference
3D model, using the Hausdorff distance , the differences
are measured and highlighted using a color palette. F or a
correct estimation of the differences size, we have to take
into account the RMS of the 3D model.
Removing the background from the images leads to the
obtaining of a point cloud with less noise and it doesn't
affect the process of their correlation. F or further
improvement in the noise removal and for reducing post –
processing workload, it is proposed the solution of
eliminating improper pixels based on the color code,
because the texture of the stone is somewhat uniform and
differs from the background.
Although the point cloud density decreases significantly
at the same time with decreasing the resolution or
eliminating the background, the final result enables the
determination of the block stone size (length, width,
height).

Journal of Geodesy, Cartography and Cadastre
36
After comparing t he point c loud automatically generated
based on digital images with the reference model, namely
the mesh surface created based on the TLS point cloud ,
the differences are in range of -0.1 cm and 1.8 cm for
most of the points, and for a small number of points
represe nting 4.6% of the total number of points , the
differences are situated outside the range , concluding that
these points contain noise.
In the case of the points situated on the block stone sides ,
calculated using the images for which the rays are
parallel , the differences are bigger than in the case of the
points situated on the block stone top and front parts,
calculated using the images for which the rays are
convergent .
The block stone 3D reconstruction process based on
digital images , satisfies the acc uracy required by the
project, of a few centimeters , so it can be concluded that
this technology can be successfully used for this intended
purpose.

Acknowledgment
This work was supported by Technical University
Vienna, the practical tests being made in t heir
laboratories. A part of this research has been done with
the help of Philipp GLIRA (Department of Geodesy and
Geoinformation); the main ideas of this paper were
developed during an experience exchange at TUWIEN in
March -June 2014.

References

[1] Aspert N., Santa -Cruz D., Ebrahimi,T. Mesh:
measuring errors between surfaces using the
Hausdorff distance. In Proc. of the IEEE International
Conference in Multimedia and Expo (ICME) , vol. 1,
pp. 705 -708, Lausanne, Swissa, 2002 .
[2] Beall C., Dellaert F., Indelman V., Roberts R.,
Incremental Light Bundle Adjustment, Georgia
Institute of Technology , 2012
[3] Cignoni P., Rocchini C., Scopigno R., Metro:
Measuring error on simplified surfaces. Computer
Graphics Forum , 17 (2), pp. 167–174, 1998 .
[4] Krajňák M., Pukanská K. and Bartoš K., Application
of digital photogrammetry in the process of
documentation of archaeological artefacts , Acta
Montanistica Slovaca, Vol.4, 2011.
[5] Kraus K. , Photogrammetry, Volumes I and II.
Dummler, 1997.
[6] Kutulakos K. N. and Vallino J. R.. Calibration -free augmented reality. IEEE Transactions on
Visualization and Computer Graphics, 4(1):1 –20,
1998
[7] Linder Wilfried, Digital Photogrammetry, Springer,
ISBN -10 3-540-29152 -0, 2006
[8] Luhmann T., Robson S., Kyle S., Boehm J., Close –
Range Photogrammetry and 3D Imaging, ISBN 978 –
3-11-0300269 -1, Germany ,2013
[9] Luhmann T., Robson S., Kyle S., Harley I., Close –
Range Photogrammetry: Principles, techniques and
applications, ISBN 1 -870325 -50-8, UK, 2006.
[10] McGlone, J., Analytical Data -Reduction Schemes in
Non-Topographic Photogrammetry. In H. Karara, ed.
Non-Topographic Photogrammetry. 2nd ed.
American Society for Photogrammetry and Remote
Sensing. pp.37 -57, 1989.
[11] Mikhail Edward M. and Bethel James S., Introdution
to Modern photogrammetry, ISBN 0 -471-60924 -9,
USA, 2001.
[12] Oniga V. E., Chirilă C., Hausdorff distance for the
differences calculation between 3D surfaces,
RevCAD 15, pp. 193 -202, 2013.
[13] Robertson D.P. and Cipolla R., Structure from
motion, In Varga, M., editors, Practical Image
Processing and Computer Vision, John Wiley, 2009.
[14] Roy M., Foufou S., Truchetet F., Mesh comparison
using attribute deviation metric, International Journal
of Image and Graphics (IJIG), Vol. 4, Nr. 1, pp. 127 –
140, 2004.
[15] Ruther H., Smit J. and Kamamba D., A comparison
of close -range photogram metry to terrestrial laser
scanning for heritage documentation, South African
Journal of Geomatics, Vol. 1, No.2, August 2012 .
[16] Singh V. P., Singh P. and Haritashya U.K.,
Encyclopedia of snow, ice and glaciers, Springer,
ISBN 978 -90-481-2641 -5, 2011.
[17] Spetsa kis ME ., Aloimonos Y. , A multiframe
approach to visual motion perception , International
Journal of Computer Vision 6 , pp. 245-255, 1991.
[18] Triggs B., McLauchlan P., Hartley R., and
Fitzgibbon A., Bundle adjustment – A modern
synthesis. In Triggs, B., Ziss erman, A., and Szeliski,
R., (Eds.), Vision Algorithms: Theory and Practice,
LNCS, p p. 298–375. Springer Verlag , 2000.
[19] Zisserman A., Fitzgibbon A., and Cross G. , VHS to
VRML: 3D graphical models from video sequences ,
In ICMCS, p p. 51–57, 1999.
Information on
http://www.danielgm.net/cc/documentation.html

Journal of Geodesy, Cartography and Cadastre
37

Investigation on the influence of the incid ence angl e on the
refl ectorl ess distance measurement of a terrestrial laser scanner

Miriam Zámečníková, Hans Neuner, Stefan Pegritz, Robert Sonnleitner
Rece ived: April 2015 / Accepted: September 2015 / Published : December 2015
© Revista de Geodezie, Cartografie și Cadastru/ UGR

Abstrac t
Although the influence of incidence angle (IA) is one of the
known err or influence of terrestrial laser scanners (TLS), it
is not taken into account in the evaluation of TLS -data. In
this paper fundamental question is discussed, how the IA
influences the TLS -distances, it is of stochastic or of
systematic nature or of a comb ination of both. For this
purpose, a new methodology has been developed. Its
special feature is that the directly measured TLS -distances
are compared with reference distances. It is optional for
close range and for longer distances. The methodology was
realised with a time of flight laser scanner. At close range
of 3.5 to 5.2 m other error effects up to 4.4 mm are more
pronounced than the influence of IA. At the distance of
about 30 m, a systematic effect of the IA was found. The
total variation of the dist ance difference with IA is of ca.
2.0 mm. The stochastic properties of the influence of IA
could not be quantified. In future works the methodology
will be improved with respect to the obtained knowledge
and the error influence well be completely quantifie d.

Keywords
incidence angle, reflectorless distance measurement, laser
scanner

M. Zámečníková, H . Neuner, S . Pegritz, R . Sonnleitner
Vienna University of Technology
Department for Geodesy and Geoinformation
Research Group Engineering Geode sy
Gußßhausstraße 25 -27, 1040 Wien
E-mail: hans.neuner@geo.tuwien.ac.at . 1. Introduction

In general, the geometry of object surfaces is determined
from terrestrial laser scanning (TLS) measurements under
varying incidence angles (IA). In consequence, the l aser spot
is deformed so that less signal strength is reflected back in
comparison to its perpendicular alignment. The IA of the
laser can affect the reflectorless distance measurements (RL)
and thus, the TLS -data. In order to consider this influence in
the TLS -measurement’s planning as well as in the evaluation
of TLS -data and in the object modelling, the quantification of
its impact is necessary.
Previous investigations of the influence of the IA on the
distance measurement of TLS are characterized by thr ee
problems. First, it is not clear which character the error
influence has. In some studies it has been described by a
systematic measure like a correction term [1, 2, 3] and in
others by a stochastic measure like a standard deviation [4,
5]. Secondly, th e impact was assessed by indirectly derived
parameters. These can also falsify the quantified influence of
the error. For the third the impact of the IA was quantified
only at close range.
In this paper the influence of IA on the RL measurement is
quantifi ed in such a way that in the mentioned problematic
aspects are minimized. The aim is to answer the fundamental
question, whether the influence of the IA is of stochastic or
of systematic nature or of a combination of both.
A new methodology to investigate the error influence is
introduced. Instead of deriving the parameters indirectly, the
study is performed on the level of directly TLS -measured
distances which are compared with reference distances. To
investigate the error influence at greater distances, two
variants of the method have been developed for close range
and for longer distances. The method is suitable for scanning
total stations (TLS+TS).
The proposed methodology is executed with a time of flight
TLS. The realised measurement setups and measur ement
procedures are described in detail. After evaluation of the

Journal of Geodesy, Cartography and Cadastre
38
measured data, the results are analysed, evaluated and
discussed in the framework of these research issue.

2. Methodology

Our investigations of the influence of IA are based on the
direct co mparison of the reference with the TLS -distance.
The investigated TLS – distance DTLS is defined as the
distance between the zero point of TLS+TS and the
scanned point P i.

Fig. 1: Measurement setup a) for close range, b) for longer distance

The method ology consists of the following steps (Fig. 1):
1)A planar object is scanned from a standpoint of TLS+TS.
The coordinates Y TLS, X TLS, Z TLS of the point cloud are
converted into polar coordinates Hz TLS, V TLS, D TLS.
2)The point on the object P i is staked out via Hz TLS, V TLS
using the tachometric part of the instrument and signalised.
The fundamental condition must be fulfilled, that TLS and
TS work in the same coordinate system. 3)The end points of the studied distance are determined with
a theodolite measure ment system (TMS). Subsequently, the
reference distance DRef is calculated from the determined
coordinates.
The TMS consists of TLS+TS and another TS instrument
(Fig. 1a)).
For the investigation of the error influence at longer distances
e. g. 30 m it is n ot possible to use only the TMS due to the
decrease of the accuracy and the spatial limitations in the
laboratory. In this case, the object point P i is determined with
the TMS from a base (TS1 -TS2) which is located at a short
distance to the planar object (approx. 2.5 m) (Fig. 1b)). The
base points and the reference point of the scanner are
determined in a geodetic precision network.
The variant for determining the reference distance is selected
according to the a priori accuracy analysis. The reference
should be at least one order of magnitude more accurate than
the investigated distance.
Steps 2 and 3 are repeated for distances under different IA.
4)On the basis of the differences between the TMS – and
TLS-distances the character of the influence of IA will be
investigated.

3. Measurements

The study was carried out with a Leica MultiStation MS50. It
is characterized by the accuracy of the RL -distance
measurement of 2 mm + 2 ppm, a distance measurement
noise of 0.4 mm up to 10 m, 0.5 mm up to 25 m at
measure ment frequency of 62 Hz and the angular accuracy of
0.3 mgon. The spot size is 7×10 mm at 30 m.
The MS50 was used at close range (3.5 to 5.2 m) as well as at
a distance of ca. 30 m under laboratory conditions. The near
field was chosen because instruments have special behaviour
in this range. The distance of 30 m belongs to usually
measured distances at scanning of structures. Scanning was
performed with the measuring rate of 62 Hz. The resolution
was set to be larger than the diameter of the laser spot, so that
correlations between adjacent distances are avoided.
The measurement process is automated predominantly via
GeoCOM control. In the following sub -sections the
measurement setup and the measurement procedure of the
two cases of investigations are descr ibed.
a.Experiment at close range
A wooden class -board was used as a test object (Fig. 1a)). It
has dark green color, dimensions of 5 m x 1.5 m x 0.025 m
(width x height x depth) and is almost vertically fixed to the
wall. The two station -points of the TM S were placed at 3.5 m
from the object. In this measurement setup the MS50 was
simultaneously used as a theodolite within the TMS
configuration. The base between the theodolites (TLS+TS,
TS) was 3.5 m long. For the basis determination a reference
scale of 0.8 m was positioned horizontally.
Different IAs of the laser beam are obtained by the rotation

Journal of Geodesy, Cartography and Cadastre
39
of instrument’s collimation axis in horizontal and vertical
direction. In this measurement setup the TLS -distances
vary from 3.5 m at IA of 0 gon to 5.2 m at I A of 55 gon.
In the measurement procedure first preparation steps were
performed for TMS – mutual orientation of the horizontal
circle of the theodolite and base determination. The mutual
orientation was determined by collimation in two faces.
Both instrum ents are specified with the same angular
accuracy of
Hz = 0.3 mgon. The base was indirectly
determined by solving the Hansen problem. The length of
the reference scale was measured with the laser
interferometer Agilent 5530 with
ref. scale = 0.4 ppm. The
pointing accuracy to targets of the reference scale with
MS50 is 0.3 mgon and with TS 0.3 mgon at the first and
0.7 mgon at the second end point (from 10 repetitions).
Subsequently, the board was scanned in one face with a
resolution of 0.37 gon. The atmo spheric corrections were
applied to the distance measurement.
The obtained point cloud of the object was approximated
by a plain. Hence, for each point the IA was calculated as
the angle between the normal vector of the plain and the
sighting line under Hz TLS and VTLS. The IA calculated in
this way range from 0 to 55 gon. The point cloud was
divided in 5 -gon zones of IA and 7 points per zone were
selected for further study of the distance.
Each selected point was staked out, the RL distance in the
single mo de DRL was measured and the point was
signalised with a needle. Its position was determined from
Hz, V measurements performed in two faces from the two
TMS -stations. The points located in two zones were
determined twice, in order to empirically determine t he
accuracy of the stacking out and of the reference
measurement. A maximum deviation of the reference
distance of 0.4 mm was obtained by this procedure.
The stability of the stations was monitored during the
measurement process; first by collimation, seco ndly by
repeated measurements to surrounding prisms, and third by
repeated base determination.
Within a time interval of 2 months the measurements were
performed with two different TLS+TS instruments using
the same measuring setup and another measuring
arrangement with a longer base of about 7 m as well.
b.Experiment at 30 m -long distance
The test object used in this case was a granite board with
dimensions 0.40 x 0.40 x 0.03 m (width x height x depth),
that has a smooth and a rough side (Fig. 2). It was p laced
nearly vertically on a Thorlabs board and fixed laterally.
The Thorlabs board with weight of 30 kg and dimensions
of 0.60 x 0.60 x 0.06 m is sufficiently stable for the granite
board.
The different IAs were obtained by rotating the object
around its vertical axis. For this purpose, an angular scale
was used. The TLS+TS was installed on a pillar about 30 m
away from the test object. The distance between the two theodolites (TS1 and TS2) forming the TMS was 2.4 m. The
basis was placed at a distance of ca. 2 m from the object. The
three instrument stations and the surrounding 8 prisms (Ri)
mounted on consoles and pillars built the geodetic precision
network.
The measurement campaign started with the determination of
the precise network. During the entire campaign the three
instruments remained mounted in tribrachs to avoid centering
errors. Therefore, with each instrument (TLS+TS, TS1, TS2)
the elements Hz, V, D were measured to the prisms while
only Hz, V were measured to the other instruments by the
collimation in 3 sets. TS1 and TS2 have a specified angular
accuracy of 0.3 mgon, TS1 the distance accuracy of 2 mm +
2 ppm and TS2 1 mm + 1.5 ppm.
For each IA the granite board was scanned with a resolution
of 0.0212 gon in one face. Just as in close range i t was then
approximated by a plain. At each IA among all scanned
points 5 per position were selected on the basis of their
distance to the adjusted plain. Each selected point was staked
out and signalised in the HzTLS, VTLS direction. The 3D –
position of th e signalised point was determined in two faces
with TMS. The granite board was aligned in steps of 10 gons
in order to get IA between 0 and 60 gons. The influence of
the IA was studied on both sides of the granite board. At IAs
of 0, 45, 55 gons staking ou t and TMS -measurements were
realised twice, in order to quantify the accuracy.
By means of measuring 4 points on the board before and
after staking out it was verified if the position of the board
remained unchanged during the staking out and reference
measurement process (Fig. 2).
The stability of three stations was controlled by polar
measurements to prisms and the Hz, V directions
measurements between stations.

4. Post processing and results

The reference distances were determined from the highly
accurat e measurements. They meet the high precision
requirement that is necessary in order to quantify the
influence of the IA. Any systematic deviation affecting these
measurements was first analysed. Based on this assessment
the accuracy achieved for the refere nce range could be
expressed. Furthermore, reference and TLS -distances were
compared, the resulting distance differences were analysed
and conclusions were drawn.
4.1.Investigation at close range
The reference distances are determined from the coordinates
of the TLS+TS -zero point and of the selected object points.
Errors that could possibly affect the obtained reference
distance are listed in the Tab. 1. They were methodically
eliminated or quantified and their impact was evaluated.
Based on this research w e conclude that the reference
distance could be distorted systematically up to ca. 0.2 mm.

Journal of Geodesy, Cartography and Cadastre
40
Tab. 1: Error influences on the reference distance determination in
close range
Influence Impact/Elimination
Stability of the
theodolite 1.Repeated measurement of 5
prisms max. coordinate
difference of 0.5 mm – within
the accuracy of the
measurement method ;
2.Repeated collimation – emp.
 of 0.5 mgon
3.Repeated base determination
 of 0.1 mm, max. deviation of
0.3 mm
Stable stations
Axes errors,
eccentricity error s Eliminated by measurements in
two faces
Skewness of the
trunnion axis Min. impact at V directions
from 95 to 105 gon
Collimation emp.  of 0.5 mgon, max.
dev.1.2 mgon
Max. impact on the reference
distance 0.2 mm
Base determination  of 0.1 mm, max. dev. 0.3 mm
Hz, V –
Scanning/Staking out Max. dev. in Hz und V of 0.8
mgon
Max. impact on the reference
distance 0.02 mm
no influence
Intersection angle 45-58 gon
Measurement with another
configuration with doubled base
length
no influence
Staking ou t/TMS repeatability of reference
distance of0.2 mm

The a priori accuracy of the reference distance of 0.2 mm
was obtained by simulation studies. This value conforms
exactly to the empirical standard deviation of the reference
distance, obtained from two independent repeated
determinations of the reference distance in two zones.
The differences between the reference distances D TMS and
the corresponding distances in the scanning mode D TLS are
shown in Fig. 3. The illustrated differences vary
systematically with the IA. The scanned distances are up to
3.0 mm longer than D TMS in two intervals: 0 -35 and 50 -55
gon. In contrast, the distances are up to 4.4 mm shorter
within the interval 35 -50 gon. The shown systematic effect
is physically or geometrically not -explainable. It was
therefore assumed, th at the obtained effect results from a
superposition of the influence of IA with other effects in
close range.
The systematic difference between the reference and the
scanned distances was reproduced 1.5 months later with
another instrument of the same typ e using the same
configuration as well as a slightly modified configuration
with a longer base (Fig. 3 a)).

Fig. 3: Distance differences as function of the incidence angle
a) Differences between D TMS and D TLS,
b) Differences between D TMS –DRL

If the systematic part of the distance deviation is split up
using an appropriate approximating polynomial function, the
stochastic properties in each zone of IA can be quantified. In
this case, it is not relevant to express the precision as a
function of the IA.
The differences D TMS-DRL show no systematic effects. The
distance deviations are mainly in the interval of -1.0 mm to
1.5 mm, which corresponds to the manufacturer specification
(Fig. 3b)).
To explain the occurred systematic effect in TLS -distance
(Fig. 3 a)) further analysis and experiments were performed.
The conceptual connection of the investigations is:
a) Determination of the distance dependence.
b) Indication of the surface dependency.
c) Determination of the colour dependence.

Journal of Geodesy, Cartography and Cadastre
41
In the experimental se tup not only the IA varies, but also
the distances. Therefore, the differences D TMS-DTLS were
plotted as a function of distance in Fig. 4. Obvious distance
dependence in the form of a cyclic oscillation can be
noticed. However, this could not be a cyclic p hase error
because the instrument uses the time of flight method for
distance measurements. To split up the influence of the
distance a measuring arrangement with a fixed distance
(minimal distance variation) and variable IA needs to be
realised in the fut ure.

Fig. 4: Differences D TMS – DTLS as function of the distance

B) Material dependence
The RL -distances measured in the single mode showed a
good agreement with the TMS -distances (Fig. 3b)). For this
reason, the former are following used as a reference basis
for comparison. The board and parts of the adjacent white
concrete wall have been scanned.

Fig. 5: Differences D TMS – DTLS as function of the material (Abscissa
Y-coordinate, almost parallel to the board)
The distances to some points were measu red in RL -modus.
The differences between RL – and scanned distances are
shown in Fig. 5 and indicate that the systematic effect is
occurring only for the dark green board. Thus, the material
dependence is evident. It should be noticed that the board has
much lower reflectivity (8%) than thewall (90%) (empirically
determined using Kodak gray card).
C) Dependency on the colour
Another board of the same colour and of another material
consisting of a layer of glass and chipboard was examined as
in the previous e xperiment. In addition, different light
colours were applied with chalk. The systematic differences
occur only in case of dark green surfaces (Fig. 6).

Fig. 6: Differences DTMS – DTLS as function of the distance, distances
were measured to a surface of different colours

At the close range the systematic cyclic error effect
influences TLS -distances. It occurs by scanning of dark
green material.
4.2. Investigation at close range
The longer reference distances were determined in two steps.
First the coordi nates of the stations were determined by a free
adjustment of the precision network. Actual instrumental
parameters were considered, which were determined by the
method of ISO17123 -4 [7]. The accuracies obtained for the
station positions are (maximum value s)
Y = 0.03 mm,
X =
0.14 mm,
Z = 0.02 mm. Secondly, the coordinates of object
points were calculated using spatial forward intersection with
the base formed by TS1 and TS2. The reference distances
were obtained from the coordinates of the zero point of
TLS+TS and the previously determined ones of the object
points.
Errors in the network measurement, the staking out and the
TMS measurement affect the determined reference distance.
Their contribution to the accuracy of the reference distance is
analysed an d summarised in Tab. 2. The highest error
influence is due to the stakeout. In our case, if the granite

Journal of Geodesy, Cartography and Cadastre
42
board rotates around the vertical axis, stakeout uncertainty
in the horizontal direction directly affects the TLS -distance
(e. g. a lateral deviation of 1 mm causes at an IA of 60 gon
a distance error dD of 1.4 mm). This uncertainty is mainly
caused by the thickness of the cross -hair and the
magnification of the telescope. In future, the scale of the
precise network e. g. the base should be controlled wit h
high-precision measurement.

Tab. 2: Error influences on the reference distance at 30 m
Influence Impact/Elimination Precision network
Points Stability of
stations 1.Repeated measurement of 8
prisms – max. deviation in a
coordinate of 0.7 mm
2.Hz, V -measurement between
instrument stations
– max. V -deviation of 1.3 mgon
– max. Hz – deviation from the
sum of the interior angles of the
triangle (TLS+TS, TS1, TS2)
1.1 mgon;
– The individual Hz-directions
vary within an interval of 2.5
mgon for TS1 and TS2, and of
0.9 mgon for TLS+ TS This
results in a probable twisting of
the Hz -circle (TS1 1.9 mgon ,
TS2 – 1.6 mgon );
The internal geometry is
preserved.
Centering error
– instruments Instruments remain in
tribraches, Hz and V –
measurement through the
collimation
Centering error
– prisms Without removing Angle Axes errors,
eccentricity
errors Eliminated in two faces
Skewness of the
trunnion axis Object points are measured
under vertical angles of 111 –
116 gon
Network points are measured
under verti cal angles of 83 –
102gon
close to the horizon, lower
impact Distance Zero points
errors considered
Scale error potential for improvement
Atmospheric
corrections considered
Accuracy of station
coordinate max. Y=0.03 mm, X=0.14
mm, Z=0.02 mm
Staking out Hz, V –
Scanning/Staking out max. dev. 0.6 mgon, lateral
deviation of 0.3 mm, distance
deviation of 0.4 mm under IA
of 60gon
Repeatability of
staked out and with
TMS determined
distance One point was stake d-out 12
times under an IA of 55 gon,
and determined with TMS
=0.29 mm
Repeatability of
staked out and with
TMS determined
distance
Twofold determination of the
reference distances of 5 points
under IA of 0,40,45,55,60 gon
=0.05 -0.51 mm TMS Accuracy of azimuth
RTS1_TS2 =0.1 mgon
Accuracy of base =0.07 mm
Angle errors As in the network
Twisting of the Hz –
circle at TS1 and TS2 Max. difference of the reference
distance of 0.02 mm
Board stability –
before/after staking
out 4 point were measure d with
TMS before and after staking
out
max. coordinate deviation of
0.05 mm
Repeatability of the
distance
determination by
TMS
1 point signalised with the
needle once and measured 12
times by TMS
=0.01 mm

The accuracy of the reference distance is calculated in the
following way:
22
Ref Net_TMS Stack   
(1)

, where
22
Stak Stack_TMS TMS   
(2)

σNet_TMS – standard deviation of the reference distance
(TLS+TS, Pi ) derived with variance propagation law by
taking into account full covariance matrix of the network
adjustment (0.17 mm),
σStak – empirical standard deviation of the staked out
reference distance,
σStak_TMS – empirical standard deviation of the repeatedl y
staked out and with TSM determined reference distance (0.05
– 0.51 mm),
σTMS – empirical standard deviation of the once signalised
and repeatedly with TMS determined reference distance
(0.01 mm);
The accuracy of the reference distance varies between 0.18
and 0.54 mm (Tab. 3)

Journal of Geodesy, Cartography and Cadastre
43
Tab. 3: Standard deviation of the reference distance [mm]
IA [gon] 0 40 45 55 60
Rough surface 0.18 0.45 0.27
Smooth surface 0.18 0.24 0.52 0.22 0.54

The individual distance differences for both sides of the
granite board a re shown in Fig. 7. In order to suppress the
measurement noise, the distance differences per IA were
averaged. The empirical standard deviations of a distance
difference per IA reach values between 0.3 and 1.0 mm.
The standard deviations of the mean values are between 0.1
to 0.4 mm. The averaged differences between the reference
and TLS -distances at each IA are illustrated in Fig. 8.
Comparing the mean values with their standard deviations
we conclude according to the 3Sigma -rule (P = 95%) that
the deviatio ns are significant (Tab. 4).

Fig. 7: Differences DTMS – DTLS as function of the incidence angle
(repeated determination – cross, star)

Fig. 8: Mean value of differences DTMS and DTLS
per one incidence angle

Fig. 9: Mean value of received signal strength by scanning to the rough
and smooth surface of the granite board

The differences (Fig. 8) have a distance offset at IA 0 gon
and vary systematically with the IA. At the rough surface of
the granite board the TLS distance is 0.8 mm longer at IA 0
gon. This difference increases at larger IA up to 2.5 mm. The
total variation of the distance difference with IA is of 1.7
mm. The smooth surface shows a similar behavior. At an IA
of 0 gon the TLS distances are longer by 1.1 mm. At 60 gon
the difference a chieves 3.1 mm. Its total variation is of 2 mm.
The significant offset (at the rough surface – not significant at
P = 95%) in the case of IA=0 gon is surprising and needs
further investigation. This IA is ideal for the RL
measurement. The offset can be cau sed by other error
influences on the RL distance measurement such as the
reflectivity of the surface or the penetration of the laser [6].

Tab. 4: Mean value of distance differences DTMS and DTLS under a
incidence angle and its standard deviation

Journal of Geodesy, Cartography and Cadastre
44
The obtained systematic variation of the distance
differences is caused most probably by the influence of the
IA on the TLS distance measurement. Higher IA lead to
worse geometrical and physical conditions, resulting in
greater distance distortion. As in the c ase of the close range
investigation the variation of the differences is strongly
correlated with the received signal strength (Fig. 9). Both,
the distance differences and the received signal strengths
are shifted (Fig. 8, Fig. 9). They also point to the i nfluence
of the surface roughness. The TLS -distances differences to
the smooth surface are in average 0.7 mm longer than the
ones for the rough surface.
In order to quantify the stochastic properties of the
distances measured under various IA, we have assu med that
the reference distances are more precise than DTLS. Under
this condition, the systematic component should be
separated and the standard deviations calculated with
respect to the IA. However, in our experiment this basic
assumption was not met. Thu s, the stochastics of the
distances among IA is not quantified. This lack of the
presented methodology needs to be eliminated in future
works.

5. Post processing and results

In this paper, a new method for investigation of the
influence of IA on the TLS -distance measurement was
presented. It is new and unique by comparing the directly
measured TLS -distances to the reference in the areal
acquisition. It is variable for distances of different lengths
and was applied here for two ranges.
At close range of 3.5 t o 5.2 m it was found out that other
error effects are more pronounced than the influence of IA.
A systematic cyclic distance -dependent effect up to 4.4 mm
was detected at a material of dark green color with low
reflectivity. Its physical cause needs to be clarified in the
future. It has been shown that in the realised measurement
arrangement with a fixed object, the variation of the
investigated TLS -distances should be minimized or even
eliminated. As a result, the object should not be fixed but
rotatable.
At the distance of about 30 m, a systematic effect of the IA
was found. For IA of 0 up to 60 gon, the TLS distances
extend over 1.7 mm at the rough and 2 mm at the smooth
surface of the granite board. The variation of the distance
differences is closely re lated to the received signal strength.
In addition, at IA of 0 gon a distance offset of -0.8 mm for
rough surface and -1.1 mm for the smooth surface could be
detected. The stochastic properties of the error influence
could not be quantified because the ref erence distances are
too noisy. From the realised investigation it can be
concluded that the accuracy of the reference distance
should be increased, especially the uncertainty of the
staking -out should be minimized. For the determination of the stochastics the methodology could be added to repeat
scanning of the object in the identical Hz and V grid. From
the repeated distance measurements stochastic properties of
the error influence can be obtained.
The first experiences show that the developed methodology
for investigating the influence of the IA has great potential.
In future works the methodology will be improved with
respect to the above mentioned shortcomings. Only the
complete answer to our research question enables a
professionally competent handling of TLS -data.

References

[1] Kersten, T., Mechelke, K., Lindstaedt, M., Sternberg, H.
(2008): Geometric Accuracy Investigations of the Latest
Terrestrial Laser Scanning Systems. In: FIG Working
Week, Stockholm.
[2] Lindstaedt, M., Kersten, T., Mechelke, K., Grae ger, T.,
Sternberg, H. (2009): Phasen im Vergleich – Erste
Untersuchungsergebnisse der Phasenvergleichsscanner
FARO Photon und Trimble GX. In: Photogrammetrie,
Laserscanning, Optische 3D -Messtechnik – Beiträge Der
Oldenburger 3D -Tage 2009, Wichmann Verlag,
Heidelberg, pp. 53 –64.
[3] Mechelke, K., Kersten, T., Lindstaedt, M. (2007):
Comparative Investigation into the Accuracy Behaviour
of the New Generation of Terrestrial Laser Scanning
Systems. In: Optical 3 -D Measurement Techniques VIII.,
Zürich, pp. 319 –327.
[4] Gordon, B. (2008a): Zur Bestimmung von
Messunsicherheiten terrestrischer Laserscanner.
Dissertation, Technische Universität Darmstadt.
[5] Gordon, B. (2008b): Diskussion von Feldprüfverfahren
zur Messunsicherheitsbestimmung für terrestrische
Laserscanner. In: Terrestrisches Laserscanning (TLS
2008), Schriftenreihe des DVW (54), Wißner Verlag,
Augsburg, pp. 125 –142.
[6] Zámečníková, M., Wieser, A., Woschitz, H., Ressl, C.
(2014b): Influence of surface reflectivity on reflectorless
electronic distance measurement and terrestrial laser
scanning. Journal of Applied Geodesy, 8 (2014), 4, pp.
311–325.
[7] ISO17123 -4 Optics and opt ical instruments – Field
procedures for testing geodetic and surveying instruments
– Part 4: Electro -optical distance meters (EDM
measurements to reflectors)

Journal of Geodesy, Cartography and Cadastre
45
Specifics of geod etic engineering measurements for staking out high
voltage lines

Petre Iuliu Dragomir, Alexandru Moldoveanu, Tudorel Silviu Clinci, Daniela Cristiana Docan
Rece ived: April 2015 / Accepted: September 2015 / Published : December 2015
© Revista de Geodezie, Cartografie și Cadastru/ UGR

Abstrac t
In this article are presented the implications of geodetic
engineering measurements in the design, verification and
execution of high voltage power lines. Depending on the
location on the globe c an arise specific situation that
involving special solutions work. In these conditions
requires integration of interdisciplinary knowledge related to
the field of terrestrial measurements such as mathematical
cartography, geodesy and surveying engineering
measurements. In the case study is presented a special
situation due to the existence in the area of two different
coordinate systems

Keywords:
high voltage lines, staking out, coordinate systems .

Prof.univ.dr.ing. – Technical University of C ivil Engineering,
petreiuliu.dragomir @gmail.com
Master ing. – moldoveanu..a@gmail.com
Asist.univ.dr.ing – Technical University of Civil Engineering –
tudorelsilviu@yahoo.com
Lecturer, eng., PhD – Technical U niversity of Civil Engineering,
daniela.docan@utcb.ro
1. Introduction

Regarding the development of high -voltage electricity
transmission lines, appeared specific objects of study such as:
– Increasing the power flowing through a power stations
and the necess ity to establish interconnections by long –
distance high -voltage and very high -voltage transmission
lines (for example, problems with parallel operation of
electric utilities that are interconnected and operate
synchronous ly, their stability to the occurren ce of short
circuits and overvoltage, the calculation of short -circuit
currents);
– Problems of interference between high -voltage and very
high-voltage power lines and telecommunication lines (for
example, setting the allowable space and distance, protectiv e
measures against disturbance or electrocution)
These kinds of objectives are the subject of the High
Voltage Technology, which is a branch of the electrical
engineering discipline with a great importance into the
power engineering field (for example, el ectrical engineering,
power engineering).
This discipline was resulted as a necessity as soon as the
first high -voltage installations was put into operation,
having as main objectives the designing, building, testing,
using and insulation of electrical dev ices in accordance with
nominal operating voltage (for a long -term regime) and
possible overvoltage [2].
High voltage domain refers in generally to voltage
exceeding 1000 volts (1kV), which is divided into the
following cl asses of electrical insulation:
– Medium Voltage (MV, C lass A): 1 kV <Nominal voltage
<52 kV;
– High Voltage (HV, Class B): 52 kV ≤ Nominal voltage
<300 kV;
– Extra high voltage (EHV Class C) ≤ Nominal voltage ≤
750 kV 300 kV;

Journal of Geodesy, Cartography and Cadastre
46
– Ultra High Voltage (UHV Class D): Nominal voltage
≥1000 kV.
However, increasing the nominal voltage amount raises
issues such as:
– The stability of the interconnected power systems;
– The necessity of using the fascicular conductors for
power lines (in particular, for nominal voltage greater than
220 kV, in order to avoid the spontaneous corona
discharges ).
The power flowing through a power station which is usually
transmitted through overhead power lines (OPL), has
increased to several thousand megawatts (1 MW=1,000,000
W), therefore, the operation of the electrical equipment must
be under the design parameters and also, must be saf ety
(high reliability) (Popovici, D. ,Lolea, M. , 2011, p. 7).
In Romania, with the fast development of the industry has
grown also the electric power industry . Thus, in 1950
performed first 110 kV line with a length of 127 km. In only
15 years, from 1950 u ntil 1965, the national electricity
transmission system has been summed up 5260 km of 110
kV lines, 663 km of 220 kV lines and 576 km 400 kV lines.
In 2012, the electric power transmission network in
Romania, managed and operated by C.N. Transelectrica
S.A., which is the National Company for Power
Transmission, included substations and electric lines with a
rated voltage higher than 110 kV distributed as follows [2]:
– 79 power transforming stations of which: a750 kV
station; 36 stations of 400 kV; 42 station s of 220 kV;
– 8931.6 kilometers of overhead lines of which: 154.6 km
of 750 kV Overhead Line (OHL), 4703.7 km 400 kV OHL;
4035.2 km of 220 kV OHL and 38 km 110 kV OHL
(interconnected lines with neighboring countries) [4].
In the overhead high -voltage lines, the geodetic engineering
measurements precede, accompany and follow construction
process. The importance of the engineering measurements
importance increase as far as automatic equipment and
methods are used in the building process and also the
measuremen t technology is improving.

2. Conditions for clearways, protection zones
and safety zones for overhead lines

The engineering surveys performed for the planning and
design OHL documentation , and stake out engineering
works, must take into account the condit ions for right -of-
way width (ROW), protection and safety zones of the
overhead lines . Essentially , the following elements should
be considered [3]:
Right -of-way width (ROW) illustrated in figure 1,
symmetrical by the centerline , is given by :
(1)
where :
–
is the maximum width of footprint transmission
towers (m);
–
is maximum length of leg members anchor or guy
wires or tensile anchors ;
–
is maximum conductor sag , under the high wind
conditions , calculated for the biggest distance between
two towers (span length expressed in meters );
–
is maximum conductors tilt under the wind pressure
(o);.
–
is horizontal safety clearance zone , from the live
conductors to its maximum deviation (m).

Fig. 1 Right -of-way width (ROW) .

Size of sites for structures for tower foundations need to be
permanently occupied during the existence of the overhead
line and areas for construction and tower assembly,
refurbishment work, maintenance equipment, ROW
clearances, staging areas, and require access road . Those
cleared areas are established according to: voltage level ,
tower design, footprint transmission towers , platforms for
conductor pulling and platforms for OHL maintenance
works .

3. Precision requirem ents

The accuracy of engineering surveying influences the
resulted conductor gauges, span length , and conductor sag.
This affects the forces applied on the structure of the
overhead line (towers, conductors, insulation, and
foundations).
The current regul ations specify that positional accuracy of
the details should be Δx, Δy, Δ h ≤ +/- 15 cm.
Accuracy of longitudinal profiles for the final
documentation should be Dx Δy Δ h ≤ +/ – 0.3 mm .
The accuracy of the horizontal angle between two
alignments should be:
(2)
where :
is angle measurement tolerance (seconds );
50cc is standard deviation for one direction is 50cc;
n is the number of measured directions .

Journal of Geodesy, Cartography and Cadastre
47
The measurement accuracy of the distances between poles is
given by :
(3)
where:
is the allowable t olerance for the observed distance
between the centers of the two consecutive towers
having a slope "p" up to 10G, and L is the horizontal distance
in meters .
For the slope of line greater than 10G, tolerance T is
increas ing with:
• 29% for slopes between 10G and 20G
• 58% for slopes between 20G and 30G
• 87% for slopes between 30G and 40G
The precision requirements for towers refer to:
– Linearity of the tower structures measured between two
towe r center …; the allowable tolerance should be:
Δ L ≤ L / 100
– Allowable vertical deviation is 0.2% of tower height.
– Crossarms deviation from the right angle; they must not
exceed 0 ,5% of the horizontal length of the crossarms from
tower center .
-Other specific requirements regarding the tower struct ure
planning (relative positioning of conductors, insulators) and
tower assembly.

4. Specific engineering Geodetic Measurements

The site surveying used for planning and designing the OHL
and the stake out engineering works in accordance with the
dimensioned plans are relay on the g eodetic support
network.
Before addressing the steps involved in achieving geodetic
support network, especially for the cross -border overhead
power systems ( OHLs), is necessary to carry out the
following tasks :
– Analysis of the ex isting geodetic points in the area of
interest;
– The study of the correspondence between the Reference
and Coordinates System of the geodetic points in the area
and the designed centers of the OHL tower;
– The coordinate transformations , if is necessar y.
There may be situations where there is not a single reference
and coordinates system used, but there are two different
reference and coordinates systems (an old one and a new
one intended to be used).
In these situations, it is necessary to analyze the reference
and coordinates systems and carrying out the necessary
transformation.
The geodetic support network design should take into
account the GNSS technology for geodetic point position ing
and total stations for the site surveying and stake out the
OHL structure .
The topographic documentation consists into a plan layout
that represents the line in a large scale and the corresponding longitudinal profiles that respect the precision requirements
mentioned previously.
Setting plan scale is based on the pr ecision requirements,
taking as a main criterion for determining the scale, the
precision representation of planimetry [ 6], according to:
(4)
where:
pl is standard deviation of the well -defined point position
on a map;
d is standard deviation of the well -defined point position
on the ground;
n is scale map denominator .

5. Case study

The case study presents a project conducted in Amman city,
the capital of Jordan. The project refers mainly to geodetic
engineering measurements and stake -out works for a
functional overhead power line of 132 kV .
Analysis of the geodetic network situation concluded there
are two coordinate reference systems: System Cassini –
Soldner and Jordan Transverse Mercator coordinates and
was identified two geode tic points in Cassini -Soldner
system. By contrast with Cassini projection, where the
reference surface is a sphere, Cassini -Soldner reference
surface is an ellipsoid . Cassini -Soldner projection is a
cylindrical projection square cross, fairness correspondi ng
projection cylinder tangent to the meridian to its
environment in the region. This is not a projection that
preserves angles unchanged, but is commonly used in the
last century to represent areas that have large expanses along
latitude. Scale after cent ral meridian is μm = 1, scale on all
verticals is μ1 = 1. Cassini -Soldner projection is more
advantageous to regions located along the central meridian,
this projection underpinning the geodetic coordinates
Rectangular – Soldner (considered by the sphere) which are
well known in geodesy. Cassini -Soldner coordinate system
is used in Jordan in both the cadastre and the special
engineering works. Cassini -Soldner projection is a
cylindrical whwre the cylinder is tangent to the sphere of the
central meridian. Small circle parallel to the central meridian
forms the north -south extent. A cylinder is designed around
the globe and is tangent along the central meridian. Due to
its similarity with Transversal Mercator projection, is the
tendency to replcenent Cassini -Soldner projection with
Transversal Mercator projection, as happened in many
countries in the Middle East. In Jordan, is intended to
replace the Cassini -Soldner system with Jordan Transverse
Mercator system for both cadastral and topographic works.
The proj ection shows a central meridian, along which the
scale remains unchanged, all other meridians and parallels
are curved , scale distortion increases rapidly with increasing

Journal of Geodesy, Cartography and Cadastre
48
distance from the central meridian [1].
Coordinate transformation from one reference system to
another can be achieved by following the steps shown
schematically in the figure below:

Fig. 2 Indirect and direct transformation of geographic coordinates [7]

Formulas for calculating latitude and longitude of the
ellipsoid by plane coordina tes north and east are [8]:

(5)

(6)
where:

(7)

(8)

is point by the central meridian with the same coordinate
point whose coordinates north must be determined and can
be calculated from the relationship:

(9)
where:
(10)
(11)
(12)
(13)
(14)
Transition by geodetic coordinates and ellipsoidal height
(B,L,HE) to Cartesian coordinates (X,Y,Z) can do the
following steps.
Considering a point P located on the surface at a height
above the ellipsoid HE, the position of this point can be
expressed through geodetic coordinates (B,L,HE) or by
Cartesian coordinates ( X,Y,Z).

Fig. 3 Transition by geodetic coordin ates and ellipsoidal height to
Cartesian coordinates

P0 is projection of point P on the reference ellipsoid ;
rp is radius of the para llel passing through the point;
N is normal sea ;
M is meridian radius of cu rvature ;
W is a function of latitude ;
WaN
(15)

321
We aM
(16)
B e W2 2sin 1
(17)
B N rp cos
(18)
NBY XHEcos2 2
(19)
The components of the position vector of a point P0 located
on the reference ellipsoid:
BabL BL B
Nr
sinsin coscos cos
20




(20)

The compone nts of the position vector of a point called “i”
located on earth surface is gived by :


 iE
i ii iE
i ii iE
i i
iii
i
B H e NL B H NL B H N
zyx
r
sin 1sin coscos cos
2

(21)

To use GNSS technology is necessary to t ransform
coordinates from the Clarke ellipsoid used as reference
surface for Cassini -Soldner projection at coordinates B,L,HE
in WGS84 system. 3D mathematical model of

Journal of Geodesy, Cartography and Cadastre
49
transformation with seven parameters has the following
form:





 









'''
,,
000
ZYX
Rm
ZYX
ZYX

(22)
'0 XRm X X 
(23) where:
X0 is translation between the origins of the two systems;
M is scale factor in ppm;
R is the orthogonal rotation matrix (R-1=RT);
 is rotation angles about X’,Y’,Z’ axes
Following these calculations are obtai ned XYZ Cartesian
coordinates on WGS84 ellipsoid. To shift the obtained
coordinates from the WGS84 ellipsoid to Hayford ellipsoid
must to add the coordinate increases corresponding to shifts
the origins of the two reference systems [1].

(24)

where:
-Cartesian coordinates on WGS84 ellipsoid;
– Cartesian coordinates on Hayford ellipsoid;
– origins shifts between the two systems.
According to prof. Steven H. Savage from the University of
Arizon a, the three transformation parameters have the
following values: X= -86m, Y= -98m, Z= -119m [2].
Conversion from Cartesian coordinates (X, Y, Z) to geodetic
ellipsoidal coordinates and ellipsoidal height (B, L, HE) is
achieved by means of the following formulas:









1
2
2 21 arctanE
i ii
i ii
iH NNe
y xZB
(25)
ii
ixyL arctan
(26)

i
ii i E
i NBy xH cos2 2
(27)

Fig. 4 Jordan Transverse Mercator projection Jordan Transverse Mercator (JTM) projection is a system
created by Jordan Royal Geographic Cent re (RGJC).
Jordan Transverse Mercator projection (JTM) is a geodetic
datum which uses the International Hayford as an Ellipsoid.
This projection is with 6° zones, central meridian of 37° and
scale factor in the central meridian of 0.9998 [9]. The
National Geodetic Network is highly accurate (Doppler
based geodetic network). Transformation parameters are not
provided by the government authorities, but coordinates N
and E in Jordan Transverse Mercator system can be
determined by the following relations [8]:
(28)

(29)
where FE is false east and EN is false north:
Values of the coefficie nts can be calculated using the
equation:

(30)

(31)

(32)

(33)

where M is the distance along the meridian from the equator
to latitude φ and can be calculated from the relationship:

(34)

where φ is the angle in radians and
is the value of M
calculated for u p
.
Points with known coordinates are in Cassini -Soldner
coordinate system. Coordinate transformation in the new
JTM was performed using software provided by the
Jordanian government authorities. The software has
implemented a number of transf ormation parameters whose
values are not made public and allow s the coordinate
transformation between the Jordan Transverse Mercator ,
Universal Transverse Mercator and Cassini -Soldner systems .

Journal of Geodesy, Cartography and Cadastre
50

Fig. 5 Coordinate transformation

Stake -out works necess ary for building OLE towers involv es
the sitting on the ground of the tower center, foundation
footings and effective foundation platform in accordance
with the coordinates indicated in the project.
For thickening of geodetic network, there were used the
following instruments: total station Leica TC Series (R)
407® with angle accuracy of 7’’and standard deviation ±
(2mm + 2ppm) for distance measurement (IR: Reflector
mode) and Leica GPS 1200+ Smart Rover with RX1210
controller having 3mm + 0.5ppm horizontal accuracy and
6mm + 0.5ppm vertical accuracy for static mode working
and 10mm + 1ppm horizontal accuracy and 20mm + 1ppm
vertical accuracy for RTK /Kinematic mode [5].
All these instruments cover the stakeout precision required
for the component of the power transmission system.
For every tower platform it is necessary to ha ve four -leg
tower to same elevation and for the cases where the four -legs
are at different elevations, the stake -out will be done
according to the project.
For each tower were made diago nal profiles who showing
the terrain model necessary for positioning of the towers.

Fig. 6 Diagonal section profile

Fig. 7 Diagonal p rofile for tower number 8

Journal of Geodesy, Cartography and Cadastre
51
Tabel 1: Coordinates tower platform 1B

Nr. pct. Y [m] X[m]
1 537192.496 385180.646
2 5372 14.762 385192.032
3 537203.574 385214.909
4 537180.499 385203.110

Nr. pct.
Y [m] X [m]
Nr. pct.
Y [m] X [m]
1 537203.385 385209.843 1 537186.453 385201.185
2 537202.497 385211.579 2 537185.565 385202.921
3 537200.761 385210.691 3 537183.829 385202.033
4 537201.649 385208.955 4 537184.717 385200.297
Nr. pct.
Y [m] X [m]
Nr. pct.
Y [m] X [m]
1 537211.431 385193.109 1 537194.997 385184.705
2 537210.703 385194.533 2 537194.269 385186.130
3 537209.278 385193.805 3 537192.844 38518 5.401
4 537210.007 385192.380 4 537193.573 385183.977
Fig. 8 Stake out sketch from tower 1B

6. Conclusions

In general, site surveying and stake out engineering works
for the overhead lines involve well known procedures.
However, particular situations r equire thoroughgoing studies
and original solutions for the engineering surveys.
The paper presents the main types of works involved in
developing OHL , highlighting the special situations that
should be solved. Therefore, the case study was analyzed the ma p projections
underlying coordinates and reference systems that can be
found in Jordan, such as: Cassini -Soldner Reference and
Coordinates System and Jordan Transverse Mercator
Reference and Coordinates System, and transformation
methods between two refer ence systems based on different
datums. It has been shown a worked example of
transformation using a software application developed by
Jordanian officials. Also, there were presented all other
engineering works performed such as: planning and design
the OHL and stake out the projected components .

References

[6] Clifford J. M., et. al. – Grids  Datums – Hasemite
Kingdom of Jordan, Photogrammetric Engineering 
Remote Sensing, December 2006, pp. 1318;
[7] Popovici , Diana ; Lolea , Mari , Curs universitar pentru
uzul studenților , Facultatea de Energetica, Universitatea
din Oradea, pp.6 -7, 2011;
[8] Normativ pentru construcția liniilor aeriene de energie
electrică cu tensiuni peste 1000 V – NTE 003/04/00 –
ISPE București, 2014;
[9] Transelectrica, Power Tranmission Grid (PTG)
Perspective Plan Period 2008 -2012 and 2017 as a guide,
p. 20, 2009;
[10] Moldoveanu , Alexandru – Proiectarea măsurătorilor
geodezice inginerești pentru realizarea LEA 132 kv –
AMMAN SOUTH to BAY ADER Al FUHEIS –lucrare
de disertație , UTCB, 2014 ;
[11] Dragomir, Petre Iu liu – Bazele măsurătorilor
inginerești –CONSPRESS, 2009;
[12] Jonathan Iliffe and Roger Lott; Datums and Map
projections 2nd Edition ; Whittles Publishing, pp.39 -89;
2012;
[13] International Association of Oil  Gas Producers –
Surveying and Positioning Guidance Not e, Number 7,
part 2, May 2009, pp.37 -39;
[14] Twinning Project Fiche – Reduce Discrepancies
between the Physical Reality and the Graphical
Cadastral Information in Jordan for the Department of
Lands and , pp.6 -7.

Journal of Geodesy, Cartography and Cadastre
52

The author's name : Ionut MAICAN
Host University : “GHEORGHE ASACHI” TECHNICAL UNIVERSITY OF IASI
Title of PhD thesis : Contributions in obtaining a geographic information system regarding the
archaeological area of a region
Scientific coordinator: Ph. D. Eng. Gheorghe NISTOR
Date public presentation : 29.04.2015

Summary of PhD thesis :

The tool that provides the proper capabilities and environment for a geo -database, that allows to store a
large amount of both graphical and textual data, providing optimum conditions for bot h inventorying and highlighting
the elements and information by separate data sets and developing multiple correlations and relations between their
common elements, thereof is the successful Geographic Information Systems.
Through this paper, we wish to e mphasize the role of Geographical Information Systems in the management of
archaeological research and in the preservation of archaeological heritage following these researches.
The first chapter highlights the importance and the purpose of topography with in the archaeological
investigations. Also, there were brought to the fore the main stages and procedures of archaeological research and the
main legislative references related heritage issues and more.
The definition of coordinate and reference systems, m ap projections aspects, issues related to coordinates’
conversions and transformations, have been the subject of chapter two .
The third chapter comprises aspects of the national geodetic network, as the decisive factor in determining
position by using the positioning technology through artificial satellites in a uniform manner all over Europe,
respecting the European standards.
Data collection technologies represented by positioning through GPS technology, data collection technologies
by means of the total station and through the integrated total stations, the data collection methods using the laser –
scanner and the digital photo -grammetry method, as well as the methods and techniques for making the digital
topographic plans, have been the subject of chapter four.
The theoretical aspects related to the implementation of an informational system and a geographical database
were described in chapter five . Thus, t003.
Here were presented the main specific definitions, a brief history of the evolution of geographic al information
systems, the interdisciplinary qualities and the areas of applicability of a GIS, the GIS functions and data models, geo –
coding and topology aspects, as well as the role of Geographical Information Systems in the management of
archaeological research and disclosing the results in GIS .
In the case study presented in chapter six, we have highlighted, phase by phase, the achievement of the
geospatial database at the preventive archaeological excavatio ns within the archaeological “Furcilor Hill” site, of the
Recea -Monolit point, Alba County. Thus, we have started from the location, the investigation of the existing
documentation on the site studied, carrying surveying – data processing and storage, and ended up with the
achievement of different conversions to adapt into GIS, data registration and management, developing the terrain’s
numerical model, efficient exploitation of informational content, as well as the elaboration of cartographic material in
GIS and the results dissemination.
The conclusions drawn from the research are grouped into: general conclusions, personal contribution in the
context of archaeological research within the site under study, and the future directions of development that focus on
the types of analysis that can be run within a GIS project, preserving the archaeological character of the information
gathered during the research.
Finally, there were presented the main references in the special literature and the electronic sources that were used in
writing the theoretical part of this work.