it is rather difficult to establish the nessecary test fields, because of the required 
elevation differences. 
In the Grimeton tests reported in [13] the camera constant and the principal 
points is determined, but it is nessecary to know the position of the exterior pro 
jection center in the co-ordinate system of the test field. The camera is located 
in a high tower and the external circumstances are changed artificially (cooling, 
filters). As the camera is not in an aeroplane, the test cannot be regarded as 
being done under operational conditions. 
In some cases there are difficulties to define the exterior projection center 
physically. As the position is determined from measurements, there also are 
errors introduced. 
8.4.1. Proposed Pest Fields. 
At Nacka, south-east of Stockholm, there are three rather high broadcasting 
antennas in a limited area. Two of them are some 190 m and the third 300 m 
high. Suppose we had targets on the antennas and on the surrounding ground 
that gave an image as in Fig. 20, then the corresponding weight and correlation 
numbers would be those in Table 14. If the standard error of unit weight of the 
image co-ordinates is 5 i/m, then the camera constant has a standard error of 
7 /<m and the co-ordinates of the principal point has 7 //m and 9 ,«m stan 
dard errors respectively. To obtain these values the ground co-ordinates must be 
very accurate. If we start by saying, that the standard error of the ground co 
ordinates has to be smaller than 1 //m in the scale of the negative, this means 
Fig. 20. 
Proposed test field in Nacka. The flight direction is chosen so that the antenna points fall in 
the corners of the picture.The heights of the antenna points are assumed to be 200 m, 200 m 
and 300 m respectively. The flying height 500 m. Wide angle camera. The corresponding 
weight and correlation numbers are given in Table 14. 
67