Visual Form pp 565-581 | Cite as

Forms Structure Form at Ever “Higher” and “Lower” Levels

  • Leonard Uhr


The central purpose of perception is to recognize and describe: From 2-dimensional sensed images, which 3-dimensional objects, moving about in the 4th dimension of time, are “out there?” Almost always form (shape, structure) almost certainly contains the most important information for identifying objects. Even small fragments of form are usually enough; but images with absolutely no shape information are almost always enigmas — until we structure them, much as we do Rorschach blobs, with our own internal forms.


Input Layer Basic Module Subsequent Layer Computer Vision System Output Window 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Honavar, V. and Uhr, L., 1989, Experimental results indicate that generation, local receptive fields and global convergence improve perceptual learning in connectionist networks, Proc. Int. Joint Conf. on Art. Intell.Google Scholar
  2. Honavar, V. and Uhr, L., 1990, Brain-structured connectionist networks that perceive and learn, Connection Science, 1:139–159.CrossRefGoogle Scholar
  3. Li, Z. N. and Uhr, L., 1987, Pyramid vision using key features to integrate image-driven bottom-up and model-driven top-down processes, IEEE Trans. Systems, Man and Cyber., 16:250–262.CrossRefGoogle Scholar
  4. Rosenfeld, A. (Ed.), 1984, “Multiresolution Image Processing and Analysis”, Berlin: Springer-Verlag.zbMATHGoogle Scholar
  5. Tanimoto, S. L. and Klinger, A., (Eds.), 1980, “Structured Computer Vision: Machine Perception Through Hierarchical Computation Structures”, New York: Academic Press.Google Scholar
  6. Uhr, L., 1978, “Recognition cones” and some test results. in: “Computer Vision Systems”, A. Hanson and E. Riseman (Eds.), New York: Academic Press, 363–372.Google Scholar
  7. Uhr, L., 1987. Highly parallel, hierarchical, recognition cone perceptual structures, in: “Parallel Computer Vision”, L. Uhr (Ed.), New York: Academic Press, 249–292.Google Scholar
  8. Uhr, L., 1990, Increasing the power of connectionist networks, by using structure, information, learning, Connection Science, 2, 179–193.CrossRefGoogle Scholar
  9. Uhr, L., 1991, Cycling logarithmically converging nets that flow information to behave (perceive) and learn. in: “Neural Networks for Human and Machine Perception,” Harry Wechsler (Ed.).Google Scholar
  10. Uhr, L., Information-Transforming Structures (ITS): Body/Mind Models of Perception and Cognition, in preparation.Google Scholar
  11. Uhr, L. and Douglass, R., 1979, A parallel-serial recognition cone system for perception, Pattern Recognition, 11:29–40.zbMATHCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 1992

Authors and Affiliations

  • Leonard Uhr
    • 1
  1. 1.Department of Computer SciencesUniversity of WisconsinMadisonUSA

Personalised recommendations