3.2 Image-variant interior orientation and correction grid 
Due to high variation in image-variant parameters, particularly 
in y-direction, and especially for the CLC3 caused by careless 
handling, high peaks in positive and negative direction could be 
expected. This is figured in the adjustment results. Fig. 4 
illustrates the variation of image-variant interior parameters in 
chronological order of exposures (exemplary forCLC1). 
  
variation in principal distance 
  
  
  
  
  
deviation in mm 
  
  
  
  
  
  
Images 
  
  
variation in principal point (xh) 
deviation in mm 
eo 
    
Images 
  
  
variation in principal point (yh) 
  
o 
f o 
> 
an 
  
  
  
  
  
  
  
deviation in mm 
37 
4 
53 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
Images 
  
Figure 4: Variation in principal distance and principal point 
The behaviour of both cameras, the Fuji S1 Pro and the Kodak 
DCS 460, is similar. By rolling the camera around the optical 
axis with —90? negative deviations occur in x-direction while 
they are positive in y-direction. Rolling through +90° yields 
negative values for y-direction and reverse positive in x- 
direction. High positive deviations in y-direction arise from 
rolling through 180°. 
Shortis et al. (1998) assumed that the position of the principal 
point varies in a specific manner with respect to the imaging 
direction. Likewise shown in Fig. 5b for the CLC2 the position 
varies around a coordinate-origin in different directions. This 
assumption does not occur for each data set as it is dependent on 
the stability and sensitivity of the used camera. 
The principal point mainly varies like assumed. Different spread 
effects are visible while the main variation of the position of 
one viewing direction is similar and varies in an average of 
+20pm. 
  
   
  
  
  
  
  
  
  
  
  
  
  
  
a position of principal point - CLC 1 
0.04 
0.03 
0.02 
0.01 
€ 0 
-0.01 
-0.02 
-0.03 
-0.04 4 3 i 1 
-0.04 -0.03 -0.02 -001 0 
xh 
b 
position of principal point - CLC 2 
0.04 4 
0.03 | 
0.02 4 
0.01 
= 0 
0.01 
-0.02 4 
-0.03 
-0.04 T 7 i 
-0.04 -0.03 -002 -001 0 001 002 003 004 
xh 
C 
position of principal point - CLC 3 
0.04 
0.03 
0.02 
0.01 
€ 0 
-0.01 
-0.02 4 
-0.03 
-0.04 * + + 1 
-0.04 -0.03 -002 -001 0 001 002 003 0.04 
xh 
  
  
  
Figure 5: location-plot of image-variant principal point 
The values of principal distance and principal point of each 
image of one data set varies within X33um where the maximum 
and minimum values are separately listed in Table 3. 
  
  
  
  
  
  
  
  
  
PCC | CLC2 | CLC3 
Mean values 
Ci 29.15337 29.15015 28.65453 
Xh 0.29152 0.26663 0.55286 
yh -0.05304 -0.11359 0.23434 
Maximum Minimum 
Cj 0.02522 0.01034 0.00999 
-0.01070 -0.01677 -0.01288 
Xp 0.01853 0.00896 0.0198 
-0.02181 -0.00827 -0.01063 
yh 0.03072 0.02961 0.03329 
-0.02276 -0.02563 -0.02115 
  
  
  
  
  
Table 3: Values of image-variant parameters 
The principal distance mainly varies within +10um for all sets 
of images. Having a closer look at the behaviour of the principal 
point some differences are coming up. While the principal point 
in x-directions varies within +10um as the principal distance 
does equally, the variation in y-direction is higher than 
expected. The deviations range within +20um with much more 
peaks up to +30um. The image-variant principal point for CLC3 
varies more compared to the other sets of images. 
—30—