Time Required for the Various Operations
The participants were requested to supply data on the time used in
various orthophoto operations. Since this experiment was based on only one
single stereomodel, the information available is not too significant.
However, some general conclusions on the time involved can be drawn.
The initial photogrammetric operation, including the preparation
of the control manuscript and the relative and absolute orientation
procedures, should not differ significantly for the various analog instru-
ments used; it should be approximately the same as for conventional photo-
grammetric plotting operations. An average of 1.5 hrs. was reported by the
participants for this operation. For the Gestalt Photo Mapper, which is
based on an analytical relative and absolute orientation procedure
supported by the automatic image correlation, the combined time required
for preparation and orientation was 37 minutes.
The time required for the profiling procedure depends, among
other things, upon the profile density, the terrain characteristics, the
scanning speed selected by the operator, and, for certain instruments, on
the enlargement factor used for the orthophoto production. Most instru-
ments offer the opportunity to vary the scanning speed. Expressed at the
scale of the photographs, these ranged from 1.0 mm/sec to 2.0 mm/sec.
The average scanning speed was 1.6 mm/sec. A slit length of 5 mm at
orthophoto scale and 4x enlargement would correspond with a slit length
of 1.25 mm in the original photo. Assuming a 60% longitudinal and 30%
side overlap, the average scanning time for a stereomodel would be 2 hours,
not counting the time for occasional rests which, we think, cannot be
avoided.
For instruments in which slope corrections normal to the profile
direction are applied, a larger slit size can be used, resulting in a
significant reduction of scanning time. This is demonstrated for the
Wild OR 1 (slit in photo scale 2 mm) and Zeiss GZ-1 (slit in photo scale
2.4 mm) for which the times reported for the manual profiling operation
were 1 hr. 20 min. and 1 hr. 45 min., respectively. The use of automatic
image correlation results in a further time reduction. For instance, the
time reported for the Gestalt Photo Mapper was 1 hour to produce each,
orthophoto together with simultaneously recorded and plotted height
information.
It is interesting tonote that in this experiment the time needed
to scan the model in a profiling mode is considerably shorter than the
time needed for contouring. It took an experienced operator 8 hours to
produce the pencil-drawn contours of the Ripon test area on the Wild A7
(Fig. 2) as compared to an average of approximately 3 hours reported for
the manual profiling of the same model. Obviously the question of
accuracy should also be considered when comparing both approaches.
The scanning speeds in off-line orthophoto projectors which use