REMOTE SENSING LiDAR PHOTOGRAMMETRY [622471]

REMOTE SENSING LiDAR & PHOTOGRAMMETRY
19 May 2017

SERVICES
Visual Inspections
Photo mapsDigital Terrain Models
Thermal InspectionsAerial Imagery
Volume Computations
Aerial Video Training & Consultancy

SYSTEMS
3MD4 -1000
HEF-30 (2x)Zenith (2 x)
DJI Inspire I (2x)
Asctec Falcon V8 (3x)
Trimble UX 5 HP
Cessna
Balloon (5x)
SkeyeBat

CLIENTS

UAV LiDAR vs PHOTOGRAMMETRY
5
🥊
🥊

LiDAR PRINCIPLE
6Distance = Time of travel / 2
Speed of lightTransmitter
ReceiverReflector

BATHYMETRIC LiDAR
7

LiDAR PRINCIPLE ACTIVE LIGHT
8

POSITIONING LIDAR
9

10
POSITIONING LIDAR

11
POSITIONING LIDAR

12
POSITIONING LIDAR

13
POSITIONING LIDAR
POSITION AND ORIENTATION ERRORS
ARE NOT THE SAME FOR ALL
RETURNS
PER SCAN
=> NOT CORRELATED

LiDAR ERROR SOURCES
14Sensor Position
GPS error
INS/IMU error
GPS -IMU Integration error
Angular Errors
Misalignment between LiDAR scanner and IMU (Boresight
calibration)
Lever arm Error
Incorrect positioning between GPS antenna and LiDAR sensor
LiDAR Range Error
Precision of LiDAR scanner
Divergence of Laser beam
Multipath error
Reflection on a sloping surface

LiDAR ERROR SOURCES
15Range
Between 5 mm to 20 mm
Position
With RTK or PPP Positioning between 15 mm and 50 mm
Orientation
Between 0.025 degrees and 0.15 degrees
Example Sum of all errors
Velodyne HDL 32E Scanner
Flying Height 60 meters AGL (Above Ground Level)
Range error: <= 20 mm
GNSS Positioning
Horizontal: 1 cm + 1ppm, assume 11mm
Vertical : 1.5 times horizontal = 16.5 mm
Total = √(11 mm2+ 16.5 mm2) =19.83 mm
Range and Positioning error: 20 mm + 19.83 mm = 39.83 mm
IMU accuracy Pitch and roll: 0.15°⟹60 meters Range = 60 * tan(0.15 °) = 15.7 cm
IMU accuracy Pitch and roll: 0.025°⟹60 meters Range = 60 * tan(0.015 °) = 2.62 cm
Total Error = √(15.72+ 3.92)= 16.18 cm /Total Error = √(2.622+ 3.92)= 4.7 cm

LiDAR PROJECT SCHEVENINGEN BREAKWATER

PHOTOGRAMMETRY
17

PHOTOGRAMMETRY
18

PHOTOGRAMMETRY
19

PHOTOGRAMMETRY
20
POSITION AND ORIENTATION ERRORS
ARE THE SAME FOR ALL PIXELS
PER PHOTOGRAPH
CORRELATED
NOT WITH ROLLING SHUTTER !!

Rolling Shutter and Photogrammetry
21

STEREO VIEWING
22

PHOTOGRAMMETRY

PHOTOGRAMMETRY

PHOTOGRAMMETRY ALLIGNMENT

PHOTOGRAMMETRY

PHOTOGRAMMETRY

PHOTOGRAMMETRY ACCURACIES
General ‘rules of thumb’ for photogrammetry with dense matching techniques
-Relative accuracy is influenced by resolution (GSD, Ground Sampling Distance)
-Absolute accuracy is influenced by quality of the geodetic network (i.e. ground control points)
-Absolute accuracy is influenced by the data processing methodology
-If all of the above are favorable:
-X,Y accuracy is 1 to 1.5 times the GSD
-Z accuracy is 1.5 to 2 times the GSD
-Absolute accuracy is the quality of the network + relative accuracy
Sample project Scheveningen breakwater
-Flight altitude 40 meters with Sony A7r (36 Mp and 35mm lens) => GSD = 0.7 cm
-Quality of the Ground control points assumed at 2cm X,Y and 3 cm Z
-A priori estimated error = √((1.5∗0.7)2+ 32)= 3.18 cm

PHOTOGRAMMETRY SAMPLE PROJECT

PHOTOGRAMMETRY SAMPLE PROJECT

PHOTOGRAMMETRY SAMPLE PROJECT

PHOTOGRAMMETRY ACCURACIES BREAKWATER SCHEVENINGEN
X YHeight
Level GPS DEM Dz-1 Dz-2 Absolute Dz -1 Absolute Dz -2
GCP0177542.555 457425.0125.676 5.681 5.686 0.005 0.010 0.005 0.010
GCP0277519.250 457437.8925.117 5.117 5.118 0.000 0.001 0.000 0.001
GCP0377524.464 457471.8874.607 4.623 4.62 0.016 0.013 0.016 0.013
GCP0477534.839 457515.8285.557 5.564 5.564 0.007 0.007 0.007 0.007
GCP0777482.622 457470.2474.542 4.544 4.541 0.002 -0.001 0.002 0.001
GCP0877455.233 457499.3664.525 4.533 4.523 0.008 -0.002 0.008 0.002
GCP1477326.597 457699.8244.519 4.521 4.516 0.002 -0.003 0.002 0.003
GCP2077285.905 457852.7784.511 4.509 4.511 -0.002 0.000 0.002 0.000
GCP2377283.584 457876.0504.496 4.502 4.494 0.006 -0.002 0.006 0.002
Average 0.005 0.003 0.005 0.004
STDEV 0.005 0.006 0.005 0.005
Dz-1 = Difference Level -GPS
Dz-2 = Difference Level -DEM

LiDAR vs PHOTOGRAMMETRY (UAV ONLY!)
LiDAR
✔Vegetation Penetration
✔Detect smaller features (i.e. power line)
✔Quicker data processing
✔No (or little) Ground control
✔Active light (better in dark/shadow areas)
✖No Picture
✖Accuracy
✖Cost
✖Weight (i.e. safety)Photogrammetry
✔Accuracy
✔Costs
✔Weight
✔Picture
✖Only map what you see
✖Longer Processing times
✖Cannot detect small features
✖Ground Control (even with RTK or PPK!)
✖Less accurate in shadow areas
CONCLUSION:
One sensor is not ’better’ than the other. Depends very much on the type of project.