hybrid system has been developed. It combines and 
integrates semantic networks for explicit knowledge 
representation with artificial neural networks which 
provide an analogous holistic object representation. 
Besides the representation formalism, the described 
system includes problem independent inference rules 
as well as a judgment based control algorithms. As 
an example for its abilities and efficiency for the 
interpretation of complex scenes, a system for the 
three-dimensional reconstruction of scenes has been 
presented. Further investigations will integrate im- 
age sequences and will also emphasize the use of 
neural network learning algorithms for the training 
of symbolic semantic networks. 
Acknowledgments 
This work has been supported by the German Re- 
search Foundation (DFG) in the project SFB 360 
*Situated Artificial Communicators". Among the 
people involved in the topic described, we want to 
thank especially to Gernot Fink, Gunther Heide- 
mann, and Nils Jungclaus for their contributions on 
the development of the entire hybrid distributed sys- 
tem. 
REFERENCES 
[1] J. Aloimonos and D. Shulman. Jntegration of 
visual Modules. Academic Press, 1989. 
[2] D. Ballard and C. Brown. Computer Vision. 
Prentice Hall, New York, 1982. 
[3] C. Brown. Issues in selective perception. In 
Proceedings of the International Conference on 
Pattern Recognition, volume 1, pages 21-30, 
1992. 
[4] M. Dhome, J. T. Lapreste, G. Rives, and 
M. Richetin. Spatial localization of modelled 
objects of revolution in monocular perspective 
vision. In Proc. First European Conference on 
Computer Vision, pages 475—485, 1990. 
[5] R.O. Duda and P.E. Hart. Pattern Classifica- 
tion and Scene Analysis. J. Wiley, New York, 
1972. 
[6] N. V. Findler, editor. On the Epistemological 
Status of Semantic Networks. Academic Press, 
New York, 1979. 
"TK. S Fu. Digital 
Pattern Recognition. Springer Verlag, Berlin, 
Heidelberg, New York, 1976. 
[8] G.  Heidemann and H. Ritter. | Ob- 
jekterkennung mit neuronalen Netzen. Report 
96/2 — Situierte K"unstliche Kommunikatoren, 
SFB 360, Universität Bielefeld, 1996. 
722 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996 
[9] F. Kummert, H. Niemann, R. Prechtel, and 
G. Sagerer. Control and Explanation in a Sig- 
nal Understanding Environment. Signal Pro- 
cessing, special issue on Intelligent Systems for 
Signal and Image Understanding’, 32:111-145, 
1993. 
[10] Mengxiang Li. Camera Calibration of a Head- 
Eye System for Active Vision. In Proc. 8 Eu- 
ropean Conference on Computer Vision, vol- 
ume I, pages 543-554, ECCV’94, Stockholm, 
Sweden, May 2-6, 1994. 
[11] Song De Ma. Conics-based stereo, motion esti- 
mation, and pose determination. International 
Journal of Computer Vision, 10(1):7—25, 1993. 
[12] H. A. Mallot. Frühe Bildverarbeitung in neu- 
ronaler Architektur. In B. Radig, editor, 
Mustererkennung 91, 13. DAGM-Symposium 
München, Informatik-Fachberichte, pages 19— 
34. Springer-Verlag, Berlin, 1991. 
[13] D. Marr. Vision. Freeman, San Francisco, 
1982. 
[14] J. Mylopoulos and H. J. Levesque. An overview 
of knowledge representation. In M. Brodie, 
J. Mylopoulos, and J. V. Schmidt, editors, On 
Conceptual Modelling: Perspectives from Arti- 
ficial Intelligence, Databases and Programming 
Languages. Springer-Verlag, New York, 1983. 
[15] H. Niemann. Klassifikation von Mustern. 
Springer-Verlag, Berlin, 1983. 
[16] H. Niemann. Pattern Analysis and Under- 
standing. Springer-Verlag, Berlin, zweite edi- 
tion, 1990. 
[17] H. Niemann, G. Sagerer, S. Schrôder, and 
F. Kummert. ERNEST: A Semantic Net- 
work System for Pattern Understanding. IEEE 
Transactions on Pattern Analysis and Machine 
Intelligence, 12(9):883-905, 1990. 
[18] E. A. Patrick. Fundamentals on Pattern Recog- 
nition. Prentice Hall, Englewood Cliffs N. J., 
1972. 
[19] D.A. Pomerleau. Rapidly adapting artificial 
neural networks for autonomous navigation. In 
R.P. Lippmann, J.E. Moody, and D.S. Touret- 
zky, editors, Advances in Neural Information 
Processing Systems 3, pages 429-435. Morgan 
Kaufman Publishers, San Mateo, CA, 1991. 
[20] W. H. Press, B. P. Flannery, S. A. Teukolsky, 
and W. T. Vetterling. Numerical Recipes in C. 
Cambridge University Press, 1988. 
[21] M. M. Richter. Prinzipien der Künstlichen In- 
telligenz. Teubner Verlag, Stuttgart, 1989. 
     
  
   
  
  
  
  
   
   
  
   
  
   
   
  
  
  
   
   
    
   
  
  
  
  
  
  
  
   
  
  
  
   
  
  
   
  
   
   
   
   
  
   
  
   
   
  
  
   
   
    
   
  
  
  
   
  
  
   
   
    
[29