ır and
trated
ention
can be
ly end
ne are
gtoa
hlight
which
x
wo
CONCLUSIONS
A contour map implicitly presents the topological relations that
exists between contours. By building the contour tree, the
adjacency of contours becomes obvious. Whilst the technique
was demonstrated by Sircar and Cebrian (1991) to be
implementable, the success of such a system relies on having
good relief plates (source data), high resolution scanner
sufficient to discriminate between closed lines, and algorithms
to 'join' adjoining sheets (figure 7). The more morphologically
simple the region, the greater the likelihood of open contours
and ambiguity.
Whilst the directed graph is an excellent method for describing
contour topology (and thus directly supports sequential
labeling), it was found that it could not be used as a descriptive
summary of the terrain; this was because the graph does not
explicitly store any information regarding the shape, orientation
and extent of each isoline. Though the area of contour regions
can be calculated by pixel counting, mathematical description of
the shape of contours is scale dependent and complex. Thus
the encapsulation of summary information (such as sinuosity)
or associating various parts of the line with geographical
features (valley, ridge, cliff, etc.) is fraught with problems.
After labeling and identifying contour lines with unique
solutions, the system should be capable of highlighting
ambiguous contours. For organizations that have large
amounts of existing paper maps and who wish to build a GIS,
this approach provides a partially automated solution. This
approach is able to establish height ordering for closed
contours, whereas for non-closed contour, the topological rules
are not applicable. The interaction between system and
operator will guide the process until all contours are labeled.
Two issues are worthy of further investigation: the human-
system interaction can be further improved by providing more
feedback from the system; and the topological rules can be
extended to inspect the contour tree for 'completeness' and
ambiguity.
REFERENCES
Boyell, R.L. and Ruston, H., 1963. Hybrid techniques for
real-time Radar simulation. In: Proc. 1963 Fall joint computer
conference, Las Vegas, Nevada.
Carstensen, L. W. and Campbell, J. B., 1991. Desktop
scanning for cartographic digitization and spatial analysis.
Photogrammetric Engineering and Remote Sensing. 57(11):
1437-1446.
Drummond, J, van Essen, R., and Boulerie, P, 1991.
Improvements to vectorizing algorithms. in: Proc. of 15th
conference and 9th general assembly, ICA, Bournemouth, UK,
Vol. 2, pp. 897-901.
Mark, D. M., 1977. Topological properties of geographic
surfaces: applications in computer cartography. in: First
international advanced study symposium on topological data
structures for geographic information systems, Dutton,
Geoffrey (ed.), Vol 2, Harvard Press.
Niemann K.O. and Howes D., 1991. Applicability of digital
terrain models for slope stability assessment, ITC journal 91-3,
pp. 127-137.
Musavi M.T., Shirvakar, M.V, Ramanathan E., and Nekovei
A.R., 1988. A vision based method to automate map
processing. Pattern Recognition, Vol. 21, No. 4, PP. 319-
326.
Petrie G. and Kennie T.J.M. (ed.), 1990. Terrain Modelling in
Surveying and Civil Engineering. Whittles Publishing,
Scotland.
269
Peuquet D. J. and Boyle A. R., 1984. Raster scanning,
processing and plotting of cartographic documents. SPAD
system, Ltd. USA
Pfaltz, J. L, 1976. Surface networks. Geographical Analysis,
Vol. 8, pp. 77-93.
Roubal, J. and Poiker, T. K., 1985. Automated contour
labeling and the contour tree. in: Proc. AutoCarto-7,
Baltimore, Maryland, pp. 472-481.
Selden, D., 1985. An approach to evaluation and benchmark
testing of cartographic data production systems. in: Proc. of
AutoCarto-7 , Baltimore, Maryland, pp. 499-509.
Sircar, J. K. and Cebrian, J. A., 1986. Application of image
processing techniques to the automated labeling of raster
digitized contour maps. in: Proc. of 2nd international
symposium on spatial data handling, Washington, D.C., pp.
171-184.
Sircar, J. K. and Cebrian J. A., 1991. An automated approach
for labeling raster digitized contour maps. Photogrammetric
Engineering and Remote Sensing, 57(7): 965-971.
Weibel R. and Heller M., Digital terrain modelling.
Geographical Information Systems: Principles and
Applications, Maguire D., Goodchild M. and Rhind D. (ed.)
Longman Scientific and Technical, UK, pp. 269-297.
Yang, J., 1990. Automatic data capture of polygon maps from
scanned data. in: Proc. FIG International Congress, Helsinki,
Finland, Comm. 5 , pp. 523-531.
Yagi, S., Kamada, K., Katsuta, K., and Okuyama, S., 1991.
Automatic digitization of 1:25,000 scale topographic maps. in:
Proc. 15th Conference and 9th General Assembly, ICA,
Bournemouth, UK, Vol. 1, pp. 286-294.
