n 
le 
d 
  
  
M,—0.043 m 0.030—0.077 m 
0.026— 0.095 m 
M,0.130 m 0.095 — 0.245 m 
0.082— 0.300 m 
Sub-commission IV:3 
Camera R.C. 5 15/23. Film. Sm,5— 11.6 microns (average). S=50 000. 
M,=0.70 m 0.49—1.26 m (5 per cent) 
0.42—1.54 m (1 per cent) 
My=0.76 m 0.54—1.38 m 
0.53—1.68 m 
M,=1.05 m 0.74—1.90 m 
0.64—2.30 m 
[t must again be strongly emphasized that these figures in principle are valid only for accidental 
errors of the fundamental operations. In the derivation of the formula systems it is assumed that the 
final coordinates and elevations are affected by influences from the same sources of errors as are the 
image coordinates and the coordinate differences (the parallaxes). Consequently, large (gross) errors 
due to lack of identification or to mistakes in the geodetic control and check point measurements etc. 
are not included in the figures above. In photogrammetry, as in all measurements, every effort must 
be made to avoid large (gross) errors as well as to determine possible systematic errors of the funda- 
mental operations. 
4. SUMMARY OF THE RESULTS OF SOME PREVIOUS TEST MEASUREMENTS 
4.1 Stereoplanigraph 
In reference [10] y-parallax measurements were performed in connection with practical tests of 
photogrammetric models. From measurements in 9 model points the mean square value of the 
y-parallaxes was found to be 0.03 mm on an average. The corresponding mean square values of the 
discrepancies in planimetry and elevation would theoretically become: 
Mr (radial)=0.49 m (0.57 m) 
M, —0.54 m (0.50 m) 
The figures within the parentheses are the corresponding results of the practical tests (the mean 
square values of the discrepancies in the check points). 
4,2 Autograph A7 
From the Oberriet tests 1954 — 1956 as reported in reference [11] interesting comparisons between 
theory and practice were made. In summary, the mean square values of the practically obtained 
discrepancies were somewhat larger than the corresponding theoretical values. 
For the three types of photographs that were tested the following averages were found: 
Image scale 1:9600: Practical results 30 per cent larger than the theoretical predictions 
Image scale 1:14000: Practical results 24 per cent larger than the theoretical predictions 
Image scale 1:21000: Agreement between practical and theoretical results. 
25