Table 1 Results for the Wild RC8 réseau camera Table 2 Results for the Zeiss RMK-A réseau 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
camera. 
No lens distortion correction No lens distortion correction 
Image Image 
deformation Self-calibration parameter set deformation Self-calibration parameter set 
correction |none a b c f k 1 m n correction [none a b c f k 1 m n 
Jámi járvi : Jami jarvi 
0 5.314.714.714.614.715.114.514.614.1 0 7.015.215.216.515.2]16.415.2]5.4|5.2 | 
1 5.2]5.0|5.0]4.6]5.115.1]4.6]4.7]4.3 1 6.0|5.4|5.4|5.8|5.5|5.9|5.4|5.5]5.4 
2 5.215.0|[4.9]4.4]15.0]5.1]4.514.3]4.2 2 6.0[5.4|[5.4|5.7|5.4|5.8]|5.4|5.4|5.3 
3 4.914.71]14.7|4.1|]4.714.714.214.3]3.8 3 5.7]5.2]15.2]15.515.215.615.115.215.0 
4 4.8|4.7|4.5|[45.1|4.71|4.7|4.2|4.3|3.8 4 $.815.215.2/5.5]15.2]5.715.215.2]5.1 
Rheidt Rheidt 
0 8.4|7.6|7.51[6.5|[7.6|8.316.616.7]|6.0 0 8.3/[6.917.4|7.7|6.9|8.0|7.0/|7.2|7.-0 
l 8.3|]7.4|7.4|]6.3]7.5|7.9]6.3]16.4|5.9 i 7.6[6.916.917.5|[6.9|7.4|6.9|6.9]6.7 
2 8.3]7.417.4]16.3|7.5]8.0]6.3]16.5]5.9 2 7.6[6.9/16.9|7.4|6.9|7.4|6.91|6.916.7 
3 8.2]7.2[7.1]6.0]7.3]7.816.216.3]5.6 3 7.4|6.616.6|7.3|6.6|7.2|6.6/[6.6]|6.4 
4 8.1]7.2]7.1]|6.0|[7.2]7.8/[6.2]6.3]5.6 4 7.4|6.716.717.4|6.7|7.3|6.7]|6.8]|6.6 
All results are root mean square vectors of the All results are root mean square vectors of the 
residuals for the ground coordinates, in image residuals for the ground coordinates, in image 
scale and in ym. scale and in ym. 
The first two steps were to be carried out to different extents, the third with all the parameter 
sets listed above. An analysis of the self-calibration parameters and their effects is to show whether 
and to what extent system calibration parameters and self-calibration parameters are identical. 
The program has not yet been completed; thus far system distortion parameters have not been 
included into the investigation, available measurements for a third testfield (Sudbury) were not yet 
used because the adjustment of the testfield survey is not yet complete, photographs for a fourth test | 
area (Seglinge) were not measured, and the planned analysis of the parameters and their effects has not 
yet been carried out. 
Tables 1 and 2 present a summary of the results obtained thus far. Several observations can be 
made on the base of these tables | 
— Image deformation correction by means of the réseau. i 
The correction of image distortion for each individual image point based on- réseau: —— M nnd 
measurements using a different number of réseau points, is more-effective for the RMK. I 
photographs, possibly as a result of the more complex deformation pattern. The more complex 
deformation pattern is also believed to be responsible for the slight loss in accuracy with | 
RMK photographs when changing from 24 to 52 réseau points for the image deformation correction | 
(Table 2, entries 3 and 4 respectively) while the addition of réseau points leads for the RC8 | | 
photographs to a slight improvement. | nn 
- Image deformation correction by means of self-calibration. E. ii 
Most parameter sets will yield for uncorrected image coordinates results. better than those 
achieved with image deformation correction but without self-calibration, with the exception of 
parameter set k in all four cases presented in Tables 1 and 2 and parameter set c when RMK 
photography was used. The parameter sets a, b and f are much more effective for RMK 
photographs and were probably developed in the course of using such photographs. 
- Overall correction by means of self-calibration. 
Parameter sets n, 1 and m perform best in the given order when both RC8 and RMK photographs are 
considered. The order appears to be determined by the effectiveness of these parameter sets 
to capture the uncorrected system distortion. Set n in particular seems capable of correcting 
the rotationally symmetrical lens distortion; while the decentering distortion present in the 
RMK photographs is not corrected as effectively. 
The investigation is continuing, and a more extensive analysis of the results will be carried out 
beyond the few observations made here. 
CONCLUSION 
The present report gives the state of a comprehensive investigation which is far from being 
completed. The presented results are the first available data and only a small fraction of the data 
Which should be available for analysis after the completion of the project. 
REFERENCES 
Bauer, H. and J. Müller, 1972: Height Accuracy Of Blocks And Bundle Adjustment With Additional 
Parameters. Presented at the 1972 ISP Congress in Ottawa. 
121