»
t area
ed in the same
where "true"
S, trees, and
ising multiple
] spacing was
wntown sites.
' of the data
. The front of
ce of the used
om the rolling
1e open areas
e terrain.
an area
Figure 8 shows a mixed area with dense distribution
of man-made features.
: Figure 8. Silver Spring, business area
Figure 9 illustrates the quality of the automatically
extracted DTM; also, to assess the need for editing,
the distribution of FOM values is shown. The
diagram covers two datasets, one over a typical
suburban subsite and one with many building
structures. As expected, there is a significant
difference in the performance of the hierarchical
correlation algorithm. Dense urban areas still pose a
difficult task for any DTM scheme.
80 -
IESuburban |
| ll Downtown |
© o
——
Frequency
N 2» O
o
0 | Om m [lm | i
void bad fair good very Figure Of Merit
good
Figure 9. Distribution of FOM
3. CONCLUSION
In a summary, after individual processing of all the
subsites in both project areas, the merged DTM
areas of Manhattan and Silver Spring resulted in a
total of 4,344,171 and 3,487,785 elevation posts,
respectively. The number of building features
extracted was 2640 for Manhattan and 1406 for
Silver Spring. The Manhattan project required 140
hours to complete, while Silver Spring was
completed within 76 hours. In addition to being
more difficult, the Manhattan site was also slightly
larger than the Silver Spring site.
Our experiences clearly show the tremendous
potential of softcopy systems for automated or semi-
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996
automated DTM and feature extraction. The
acquisition of very dense spatial data in such a large
volume is unprecedented in traditional environment.
Measuring this type of DTM information is
possible, but obviously not feasible on analytical
plotters, due to the lack of the support for the
obligatory user interactions, and furthermore,
because of the unacceptable labor requirements
necessary to manually digitize such a large number
of ground elevations. Since the automatic DTM
extraction is a batch process, the key issue in
softcopy environment is the effectiveness and the
user-friendliness of the on-line editing capabilities
and quality control tools and utilities. Our
encouraging experiences have unmistakably
demonstrated the power of softcopy systems. It has
become clear that these systems can already
efficiently compete with other existing methods. In
addition, they can deliver data in many new,
unconventional, formats, and thus, open up new
applications for the use of high-volume spatial data.
In the future, we expect substantial improvements in
the user interface, as well as continuous
development of the built-in automated processing
(Schenk and Toth, 1992). Currently, the efficient
use of these systems requires a quite considerable
amount of relevant knowledge. These changes will
ultimately speed up the proliferation of the softcopy
technology in the mapping industry.
4. REFERENCES
Helava, U.V., 1988. Object-Space Least-Squares
Coreelation, Photogrammetric Engineering and
Remote Sensing, Vol. 54, No. 6, pp. 711-714.
Schenk, T., and Toth, Ch., 1989. A PC-Based
Version of the Planicomp Analytical Plotter, Proc.
ASPRS-ACSM Annual Convention, Vol. 1, pp. 10-
18.
Schenk, T., and Toth, Ch., 1992. Conceptual Issues
of Softcopy Photogrammetric Workstations,
Photogrammetric Engineering and Remote
Sensing, Vol. 58, No. 1, pp. 101-110.
Socet Set?, 1995. Soft Copy Exploitation Tool Set,
User's Manual, Version 3.0, GDE Systems Inc.
Toth, Ch., and Schenk, T., 1990. A New Approach
for DEM Measurement with the ZEISS P-Series
Analytical Plotters, Proc. ACSM/ASPRS Fall
Convention, pp. B-145-151.
