C. Choice of methods
As already stated, the absolute accuracy of + 2-3 cm is on the borderline of what can
be achieved with reasonable means, both in terrestrial and photogrammetric surveys;
however, the use of photogrammetry promised advantages both in cost and time.
Furthermore, the organizational problems in connection with the operation of the
marshalling yard were clearly simpler, safer and more economical to solve by using
photogrammetric methods instead of conventional surveys. In particular it was possible
to organize the field work in such a way that the operation of the marshalling yard
was practically not at all delayed and the survey only a little.
Since previous practical experience with numerical-photogrammetric surveys and theo-
retical considerations indicated that the required point accuracy is really attainable,
the choice of survey method fell clearly in favour of photogrammetry. The progress
of the work and the results obtained showed convincingly that this decision was
correct.
D. The numerical-photogrammetric survey
7. Signalization
The points to be used were marked before the photo flight either with white PVC
panels on timber sleepers or white painted on steel sleepers. The size of all signals was
10 cm x 10 cm. A specially developed template was used to position the signals with
an accuracy of 1 mm in the axis of the tracks, with a significant saving of time.
(See Fig. 3.)
In the determination of the size of the signals the photograph scale of approximately
1:3000, the diameter of the floating mark in the Wild A 7 Universal-Autograph as well
as the resolving power of the Wild lens and of the photographic emulsion were taken
into account. As a consequence all points not covered by goods waggons were easily
identifiable and measurable.
2. The aerial photographs
The photographs could only be taken during a few short periods on Sundays, at
times when the marshalling yard was not in use and the through-tracks were cleared.
A postponement from a fixed day was not possible.
