From the results of these tests, the following conclusions can be drawn:
ba
Independent model aerial triangulation may be used as a substitute for analytical aerial triangulation.
2. The formulation for the adjustment of aerial triangulation should include relations between the image
coordinates.
3. Functional correlation between the image coordinates may be neglected.
23. de Masson d’Autume, G.
France
PROCESSING SYSTEMATIC ERRORS
IN AEROTRIANGULATION
The processing of systematic errors is one of the last important problems remaining to be solved in
aerotriangulation. The existence of such errors cannot be denied, and it is almost impossible to entirely
eliminate them from data no matter what precautions are taken. Their effect depends on the calculating method
adopted. It is almost nonexistent in the case of polynomial methods which are otherwise not very satisfactory.
In the case of more rigorous methods using independent models or clusters, systematic errors may introduce
extremely dangerous deformations if arrangements are not made to correct them. The method recommended
for this correction is based on an examination of remainders after compensation. Numerical, functional or
statistical filtering enables the accidental part to be separated from the systematic part of these remainders.
Different variations, some direct and others iterative, are described.
24. Meixner, H.
Germany (F.R.G.)
A UNIVERSAL COMPUTER PROGRAMME FOR
ANALYTICAL AEROTRIANGULATION
Abstract not provided.
25. Rauhala, U.
Sweden
NEW SOLUTIONS FOR FUNDAMENTAL
CALIBRATION AND TRIANGULATION PROBLEMS
A generalized theory of polynomial and harmonic curve, surface, body, time dimensioned body, etc. fitting to
measured data, yielding a strong solution for the parameter arrays with physical meaning, is given in this
paper. The generalized matrix calculus allows an effective solution of an enormous number (say 10°- 10") of
parameters, especially in some cases of problems in higher dimensions. The simple integral and derivation
rules of this theory make it of special interest in many applications.
Further, the solution of the complete camera calibration using only a 2-dimensional test field, and various new
mathematical models of camera and instrument calibrations are presented. The first experiments by the new
methods have yielded an accuracy of 1 - 2 wm for image coordinates, even with amateur cameras and acetate
film. It is evident that the new theory automatically eliminates errors from known and unknown sources in the
whole photogrammetric process, which were not eliminated by previous methods.
A most general method of performing analytical triangulation is developed by using the calibration process
in connection with the earlier theory of the author for simultaneous adjustment of photogrammetric and
geodetic observations. The first experiments, using various methods, have yielded an accuracy of 0.6 - l. 2 pm
for new points. The test camera has been the Hasselblad El Data, popularly called ‘the moon camera”.
42