ISPRS Commission III, Vol.34, Part 3A „Photogrammetric Computer Vision“, Graz, 2002 
  
  
In this case it was a question of recalibration, so that the 
starting values were very good and the corrections were very 
small. This also leads to a faster process. The final corrections 
are listed in table 3. 
  
  
  
  
  
  
  
  
  
  
  
Correction [degrees] 
R shift -0.0021 
| P shift +0.0113 
P t DU Table 3. List of result in calibration 2 
~~ " 3.3 Project Toensb 
e $^ avem. — 3. roject Toensberg 
I J 
The Toensberg project was flown in nine parallel strips (figure 
Figure 4. Profile of the calibration dataset, Above: before 6). As there is no crossing strip, there was no attempt of 
calibration. Below: after calibration solving for pitch or heading errors. Only elevation and roll shift 
were solved for. 
  
The final corrections for the calibration project was are 
presented in table 2. 
  
  
  
  
| Correction [degrees] 
| R shift +0.4028 
| P shift -0.3625 
  
  
Table 2. List of result in calibration 1 
3.2 Calibration flight 2 
In the other case, the features were large buildings with large 
roof structures (figure 5). 
  
  
  
  
  
  
  
Figure 6. The Toensberg project consists of nine parallel strips 
(left) and moderate terrain undulation (right). 
The final corrections for the Toensberg test are listed in table 
4. 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
Flightline Points Z shift R shift 
[m] [deg] 
1 1463 -0.090 +0.0589 
2 2385 -0.089 +0.0160 
3 2360 -0.020 +0.0259 
4 2192 -0.051 +0.0269 
5 2601 -0.028 +0.0194 
6 2870 -0.024 +0.0214 
7 3006 -0.046 | 40.0377 
8 2992 -0.018 | 40.0250 
9 1486 -0.050 | -0.0028 
  
  
  
  
  
= Table 4. List of result in project Toensberg 
Figure 5. System calibration flight 2, four strips flown for 
misalignment calibration. Dark areas (flat surface and roof 
structures) were used for matching. 3.4 Project Svinesund 
In this case, the ground was very flat so the roofs of the large This laser strip adjustment was made for Fotonor to calibrate 
buildings were used for matching together with four smaller the dataset and to verify the accuracy. The terrain was very 
areas with flat terrain in the corners of the overlapping area ^ undulated and the elongated area was covered by five parallel 
(dark areas in figure 5). strips and one strip across in the middle, figure 7.