. c) Height map
:ture are placed
cape model re-
al images. The
tics: roads are
hibit a continu-
p was inserted.
km x 2km) con-
texture map of
of the scene in
phic computer.
Fig. 7: Synthesized view of landscape model with continuous roads and elevated forests
7. CONCLUSION
The presented system exploits prior knowledge about the
scene to improve the realism of the model. The explicit
knowledge representation with semantic nets and rules eases
the adaptation of the knowledge base to new tasks. The ad-
vantage of the system is that the knowledge can constrain the
model parameters and select object specific surface primi-
tives. Occluded object parts and lost details due to image res-
olution are added to obtain a consistent model. Modelling
takes care of what is important for a realistic impression of a
human observer, e.g. planar roads and height steps at forest
edges.
Further work will focus on the development of the control for
interpretation and exploit multiple sensors and especially
prior interpretations of the scene represented in the german
topographic and cartographic information system ATKIS.
REFERENCES
Ackermann, F., Krzystek, P., " MATCH- T: Automatic Men-
suration of Digital Elevation Models", Proceedings of
Technical Seminar of the Sociedad Espanola de Carto-
grafia Fotogrammetria y Teledeteccion, pp.67 — 73, Bar-
celona, 12th April 1991.
Clement, V, Giraudon, G., Houzelle, S. , Sadakly, F. ,"Inter-
pretation of Remotely Sensed Images in a Context of
Multisensor Fusion Using a Multispecialist Architec-
ture”, IEEE Trans. on Geoscience and remote Sensing,
Vol. 31, No 4, July 1993.
Foresti, G.L., Murino, V., Regazzoni, C.S., Vernazza, G., "Dis-
tributed spatial reasoning for multisensory image in-
terpretation", Signal Processing, 32, pp. 217—255,
1993,
Gimmelfarb, G. L., Zalesny, A. V., "Low — Level Bayesian Seg-
mentation of Piecewise- Homogenous Noisy and
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996
Textured Images", International Journal of Imaging
Systems and technology, Vol. 3, pp. 227 —243, 1991.
Koch, R., "3—D Surface Reconstruction from Stereoscopic
Image Sequences", /CCV'95, Boston, June 1995.
Lemmens, M., "A survey on Stereo Matching Techniques",
ISPRS Conference, Kyoto, Japan, 1988.
Liedtke, C.—E., Grau, O., Growe, S., "Use of Explicit
Knowledge for the Reconstruction of 3—D Object
Geometry”. 6th International Conference Computer
Analysis of Images and Patterns CAIP'95. 6.—8. Sept.
1995 Prag.
Liedtke, C.— E., Blómer, A., "Architecture of the Knowledge
Based Configuration System for Image Analysis
CONNY”, Proceedings of the 11th IAPR International
Conference on Pattern Recognition, The Hague, Vol. I,
pp. 375—378, IEEE Computer Society Press, Sep-
tember 1992.
Matsuyama, T., Hwang, VS.—S., "SIGMA : A Knowledge—
Based Aerial Image Understanding System", Plenum
Press, New York 1990
McKeown, et al, "Rule—Based Interpretation of Aerial
Imagery”, IEEE Trans. on Pattern Analysis and
Machine Intelligence, Vol. PAMI-7, No. 5, pp.
570—585, Sept. 1985.
Mester, R., Franke, U., "Statistical model based image seg-
mentation using region growing, contour relaxation
and classification", Proc. of SPIE Conf. on Visual
Communication and Image Processing, Cambridge
Mass., Vol. 1001, November 1988.
Niemann, H., Sagerer, G., Schroder, S., Kummert, FE,
"ERNEST: A Semantic Network System for Pattern
Understanding”, IEEE Trans. on Pattern Analysis and
Machine Intelligence, Vol. 12, No. 9, pp. 883—905,
Sept. 1990.
Shapiro, S. C. (Ed.), "Encyclopedia of Artificial Intelli-
gence", John Wiley & Sons, New York, 1992.
Tónjes, R., "Realistic Landscape Modelling with High Level
of Detail", JEEE International Conference on Image
Processing, Austin, Texas, 13—16 November 1994.
