ISPRS Commission III, Vol.34, Part 3A ,,Photogrammetric Computer Vision*, Graz, 2002 
  
4. CONCLUSIONS 
The results show that the introduced parallel-sequential model 
based on Newton's equations of motion and attracting forces 
between edges may be a promising approach to real-time stereo 
processing. The algorithm gives the right disparities in most 
image points but there remain errors. Therefore, new efforts are 
necessary to enhance the quality of the approach. Some ideas 
for improvement are the following: First, far - field forces 
should be introduced. Secondly, the assumed force law (16) — 
(21) must be optimized or changed. When the right position 
Xmax(i Jj) is reached then the external forces should reduce to 
zero in order to avoid oscillations which are small but not zero 
now. 
Finally, it must be mentioned that the model can be extended to - 
the non-epipolar case introducing coordinates y(i’,j) and forces 
acting in y — direction. 
References from Journals: 
Goulermas, J.Y., Liatsis, P., 2000. A new parallel feature-based 
stereo-matching algorithm with figural continuity preservation, 
based on hybrid symbiotic genetic algorithms. Pattern 
Recognition 33, pp. 529-531 
Pajares, G., de la Cruz, J. M., 2001. Local stereo vision 
matching through the ADALINE neural network. Pattern 
Recognition Letters 22, pp. 1457-1473. 
References from Books: 
Hubel, D., 1995. Eye, Brain, and Vision. Scientific American 
Library, New York. 
Julesz, B., 1971. Foundations of Cyclopean Perception. The 
University of Chicago Press, Chicago, pp. 203-215. 
Klette, R., Schlüns, K., Koschan, A., 1998. Computer Vision. 
Springer, Singapore. 
Klette, R., Zamperoni, P., 1992. Handbuch der Operatoren für 
die Bildverarbeitung, Vieweg, Braunschweig. 
Serra, R., Zanarini, G., 1990. Complex Systems and Cognitive 
Processes. Springer, Berlin. 
References from Other Literature: 
Gimel'farb, G., 1999. Stereo Terrain Reconstruction by 
Dynamic Programming. In: Handbook of Computer Vision and 
Applications, Vol. 2, Academic Press, San Diego, pp. 505-530. 
Jahn, H., 1999a. Feature Grouping Based on Graphs and Neural 
Networks. In: Lecture Notes in Computer Science 1689, 
Springer, Berlin, pp. 568-577. 
Jahn, H., 1999b. Unsupervised Learning of Local Mean Gray 
Values for Image Pre-processing. In: Lecture Notes in Artificial 
Intelligence 1715, Springer, Berlin, pp. 64 — 74. 
A - 180 
Jahn, H., 2000a. Stereo Matching for Pushbroom Stereo 
Cameras. In: Int. Archives of Photogrammetry and Remote 
Sensing, Amsterdam, Vol. XXXIII, Part B3, pp. 436-443. 
Jahn, H., 2000b. Parallel Epipolar Stereo Matching. In: 15" Int. 
Conf. on Pattern Recognition, Barcelona, Vol. 1, pp. 402-405. 
Perona, P., T. Shiota, T., Malik, J., 1994. Anisotropic diffusion, 
in: B. ter Haar Romeny (Ed.), Geometry-Driven Diffusion in 
Computer Vision, pp. 73 - 92, Kluwer Academic Publishers, 
Dordrecht