15
localized image detail can cause an undesired
count. The improvement would be greater, except
that only information in the central part of the
scan is used for correlation, resulting in a signifi
cant time loss. The larger area provides the
operator enough of a picture to be useful, while
correlation is limited to a smaller area to obtain
a more meaningful measurement of the altitude
for the central point.
Because the effectiveness of the UAMCE oper
ational cycle remains to be verified, it is not too
useful to speculate on how much shorter the time
can become. If the equipment were improved to
a point where operator attention was not required,
the scan could be confined to the area used for
correlation. This might permit a factor of two
improvement. Faster scanning would require a
higher bandwidth and, consequently, brighter
scanners; some gain might be obtained in this
way, but it would be at the cost of poorer resolu
tion and perhaps less life from the scanner tubes.
It is anticipated that significant improvements
in the UAMCE will result from programming
modifications that will permit the density of the
measurements to adapt to the terrain. To obtain
altitude information in conformance with usual
map standards, it will be necessary to have at
least two measurements per contour interval. In
flat areas the measurement interval can be large,
allowing such areas to be covered quickly; in
areas of high relief, the density of measurement
must be very high (unless it is not important to
maintain the accuracy). It has been estimated
that the UAMCE will compile 9- by 18-inch,
100% lap diapositives in about 3 hours using (at
nominal scale) 250-micron (0.01-inch) spacing
along the profile, and 500-micron (0.02-inch)
spacing between profiles. For very flat areas this
could be changed to, say, 1000 microns between
profiles and, with modification of the computer
programs, an equal spacing along the profile; this
would permit a 9- by 18-inch, 100% lap pair to be
compiled in less than half an hour. The program
should be made to adapt to the terrain to minimize
the time required for each model, while maintain
ing the required accuracy.
The UAMCE will be improved gradually as
other new computer programs are written. These
will permit manual measurement of spot heights
with hard-copy printout, so that the data can be
added to the map. It will probably be desirable to
write a program to allow the operator to contour
manually in selected areas (as in a forest, where
the operator can use judgment to extend whatever
stereo coverage of the ground is available); the
program would print out the altitude chart on the
basis of the manually made measurements, but
print out the orthophoto on the basis of machine-
made measurements, preserving the dominant
imagery on the orthophoto. Undoubtedly, still
more programs will be found desirable to opti
mize the operation and minimize the demands on
the operator.
Of course, it would be desirable to print out
contour lines, rather than the present rotary
sequence of three gray tones. However, until
annotated contour lines (not requiring manual
touchup) can be automatically drawn, the product
would not be as useful as the present output. The
three-tone sequence is very easy to follow and
easily smoothed to obtain the final product.
The UAMCE should print orthophotos at up to
40-line/mm resolution without any noticeable
joint between the individual small areas making
up the composite output. The high resolution per
mits outputs at reduced scale, so that high quality
mosaics from contiguous stereo pairs can be made.
While orthophotos may be adequate and desirable
for large scale maps, they are not useful for small
scale maps. Continued efforts will be required to
automatically abstract the desired planimetric
detail from the orthophotos.
Many groups are working on the problem of
recognizing important details in photographs. The
compilation problem demands attention to a new
class of recognition problems (such as, in an
area under examination, can the structure of the
area be discerned from the resulting electrical
signals?). The stereo signals should make recog
nition easier (assuming the scale of the photog
raphy is such as to reveal height variation in the
features to be identified). It may be useful to be
able to place upper and lower limits for the alti
tude variations within an area rather than being
constrained to use the weighted average altitude
where the weighting favors the high information
content signals.
In the series of equipment developed thus far,
no attempt has been made to store any significant
amount of the altitude data, except through the
altitude chart. It should be realized that the com
pilation process requires that the measured height
of each elementary area be available in digital
form at some time. It would be a trivial task to
store these data on magnetic tape, making them
available for later use for many applications. For
example, the data might be used for cutting relief
models, for calculations required for cut-and-fill
operations, or for calculating the expected cover
age of a radar as a function of possible sites.
CONCLUSION
The development of the Automatic Map Com
pilation Equipment at Bunker-Ramo has been
sponsored by the U.S. Army Engineer Geodesy,
Intelligence and Mapping Research and Develop-