(= CS and CK treated in two different A7 autographs by different
operators) are shown on vectordiagrammes in Figs. 2 and 3.
The mean square value of the discrepancies in x and y as well as in
radial direction, are shown for the various models in Table 2.
The differences between in photogrammetric way determined
ground coordinates of 53 check-points determined in model A (model
AK in A7 and AV in A8) have been studied. The mean square value of
these total differences were found to be + 21 mm in x and + 17 mm
in y. The mean square value of the discrepancies of the mean photo-
grammetric coordinates (not compared with geodetic coordinates) are
consequently 11 and 9 mm respectively.
The adjustment of the photogrammetric z-coordinates has been done
With an approximate linear graphic adjustment (no electronic computer
programme available) based on the general trend of the z-errors. The
preliminary mean error in the photogrammetric elevations is roughly
+ 30 mm. A clear systematic trend, however, remains after the rough
adjustment. This trend coincides with the known minus-area in the cen-
ter of the model, which may be caused by residual lens or film distor-
tion. It should be emphazised that check-points even far out in the outer
parts of the models have been used (see Figs. 2 and 3). A normal limita-
tion of the model and a numerical adjustment including the distortion
effect would improve the mean square value of the discrepancies.
Further investigation. As soon as a programme for rigorous adjust-
ment of a stereomodel based on the residual vertical parallaxes is avail-
able such an adjustment will be done. The height adjustment will also
be studied.
Conclusion, It is quite possible to take aerial photographs from a nor-
mal speed aircraft from very low altitude under almost any weather
conditions at any time of the year and determine coordinates of signall-
ed points with an accuracy in x, y and z of about -- 30 mm. Plotting
of details and contour lines for very large scale maps is possible even in
underexposed negatives from this very low altitude if the ground sur-
face contains such contrasts and texture that the graininess of the film
will not cause an erroneous stereoscopic impression. In the described
test this was the case on the lawns of the park.
10
