ON 
; 
1 fig. 8, 
Surface 
ive dis- 
arth 
1e earth 
roup of 
at they 
he dis- 
ffect of 
of the 
and the 
vature 
524 Distortion due to lack of flatness of the pressure plate 
In connection with the analysis of the empirically determined distortion curve 
a check of the flatness of the pressure plates was also made. It turned out that 
all pressure plates had relatively large flatness errors. Deviations of up to 0.07 
mm were found. The pressure plates proved throughout to be concave, the 
concavity being relatively symmetric. If this flatness error of the pressure 
plate is conceived as an error dc in the camera constant c, the distortion cor- 
responding to this error at distance | from the image centre can be written 
l : : : 
dr, == ‘de, In fig. 10 the distortion curve corresponding to the flatness errors 
6 
of an actual pressure plate is shown. 
020 Determination of the optical distortion curve with the aid of the 
empirically determined total distortion curve 
Fig. 10 shows the result of an investigation. In the same diagram have been 
plotted both the distortion curve empirically determined by height measure- 
ment of the ice model and the distortion curves which, as mentioned above, 
can be derived from the curvature of the earth -- refraction and the lack of 
flatness of the pressure plate of the camera used. If the last mentioned two 
distortion curves are subtracted from the empirically determined total curve, 
the distortion curve of the camera is obtained. This distortion curve is in good 
agreement with the distortion stated by the camera makers. 
In fig. 11 another example is shown. In this case the ice pictures were photo- 
graphed by the Swedish Air Force from about 8,000 meters altitude (picture 
scale 1: 53,300). As shown by the figure the computed values dr, ... dr, are in 
good agreement with the distortion curve delivered by the camera makers. 
926 Discussion of the result 
The disturbances stated in the new cameras of RAK are an example of how 
several sources of error, which individually appear to be of little importance, 
result in an error that seriously affects the final result. The optical distortion 
as an isolated occurrence is insignificant in this case and may be neglected for 
most. practical purposes. It is, however, unfortunate that the distortion in this 
case has such a sign that it adds to the effects of the curvature of the earth and 
refraction. Since the pressure plates also showed a tendency to become concave, 
an error with the same sign as the other two was added in this case. As a result 
the “distortion” in the image field varied from + 20 u to — 35 yu, i.e. showed a 
total amplitude of 55 u whereas the optical distortion varied from 4- 9 u to — 6 u, 
Le. had a total amplitude of 15 pu. In stereoscopic measurement of heights in 
39