22.
plot by the operator in automatic contouring. Even then the speed of automatic con-
touring in experimental plotting is about four times the speedof anaverage operator.
To find out the maximum acceptable speed, the automatic plotting of one
contour (150m) was repeated with respective speeds of 255 sec/contour, 105 sec/
contour, 79 sec/contour, 64 sec/contour, and 58 sec/contour. In other words, on
accepting 255 sec/contour as the basic speed, the speed ratios used were 1:2.5,
1:3.2, 1:4. 0,:and 1:4.4., A striking "human" character could be detected in the
automatic contours plotted at high speed. They deviate from the correct location in
spots where terrain is flat or where there isalso the greatest ambiguity for the
human operator.
By tracing the contours in Fig. 16 on the same piece of transparent paper
the reader may easily check the repeatability of contouring at varying speeds. The
elevation errors are seldom larger than half the contour interval, or 5 meters.
In conclusion, for small-scale mapping with a contour interval of 25 m or larger,
automatic contouring at 10 times the speed of a human operator would seem to
produce satisfactory results.
Automatic Profiling
Automatic profiling is one of the simpler operations that can be per-
formed by the device. If profiles along parallel lines in one x- or y-direction are
desired, the plotting device can perform the profiling automatically. After reach-
ing the edge of the plotting table the tracing table with the scanning head is offset
by a required amount and the profiling continues automatically. The profiling can
be performed at a speed which is still faster than the top speed in automatic con-
touring. As a matter of fact, the scanning head follows the surface and the stereo-
model even at the very fast free-hand displacementof the tracingtable. The accu-
racy of profiling depends upon the profiling speed and is for average velocity
identical to the accuracy of determination of spot elevations, which has been dis-
cussed previously. It is obvious thattrees
and artificial objects have an effect on the
profile.
The present model has a simple
arrangement to draw profiles along lines
parallel to the x- and y-axis. A freely-
sliding pen is attached under 45° to the
tracing table (see Fig. 17). Any change in
the height of the tracing table results in
an identical horizontal displacement of the
pen perpendicular to the profile line. This
very simple arrangement has the draw-
back that the vertical scale of the profiles
is too small for the accurate determina-
tion of elevation differences. However,
there should be no difficulty in designinga
more elaborate mechanical profile record-
er that could permit larger scale ratios
between the vertical displacement of the
scanning head and the graphical record of
