In addition to these laboratory tests, practical experiments also were carried out. In 
three radial triangulation strips taken consecutively, for a total of 114 pictures, the 
distances between three camera references were measured. The photographic material 
was Kodak Super XX-Film, on Aerographic safety base. The pictures were developed 
on the day after the flight. The slides were made three days later. 
The deviation between consecutive pictures, as expressed by the width minus length 
difference figure, is somewhere between 20 and 30 microns. 
These results tally precisely with the laboratory tests and data given by the manu- 
facturer of the film. 
A special experiment was conducted in order finally to establish the influence of 
measured film shrinkage on the results of strip radial triangulation. 
The test grid was copied ten times on a strip of aerial film of the Gevaert-Avi-Panchro 
Hyper Rapid (Aerographic base) make. After developing and very carefully drying 
the film, slides were made from the uncut film. This was done with all due precautions, 
so that the film would not be influenced by temperature and humidity fluctuations. 
The ten slides were first measured with a precision instrument to 4- 1:5 microns. The 
vertical and horizontal parallaxes for the 6 characteristic points of a spatial model, for 
60 9; overlapping, were then determined by a comparison of these measurements with 
those taken from the original grid plate. 
Strip radial triangulation was then carried out mathematically, using data from these 
ten plates. 
The errors in the positions of the transition control points as determined by mathematical 
computation obviously are exclusively due to film shrinkage. 
The following assumptions were made for the aerotriangulation: flight altitude above 
the ground 4000 m (about 13,125 ft.), wide angle lens: f — 115 mm (4-53), picture 
size 18 x 18 cm (1-09"). 
The picture scale 1:35,000, follows naturally, with a 2,500 m base (8,200 ft.) for a 
60 9, overlap. 
Starting from these assumptions, the errors and inadequacies in the transition control 
points, due to faulty orientation elements, were computed. Furthermore, we also 
determined the errors resulting from the horizontal parallax in the transition control 
points. 
  
4205 Scale of the differences in the spatial positions of the control points 
. . . 
*AY u^ at the edges and axis of the strips 
Fig. 2 
As far as the strip radial triangulation method was concerned, we assumed that the 
altitude scale is correct at the nadir points, and that the models are adjusted for the 
nadir points.