A biologically motivated approach to early visual computations: orientation selection, texture, and optical flow

  • Steven W. Zucker
Computer Vision
Part of the Lecture Notes in Computer Science book series (LNCS, volume 301)


The problem of curve detection decomposes naturally into two stages: (i) inferring the (discrete) trace and tangent to the curve; and (ii) finding integrals through the resultant tangent field. Orientation selection is the term used by physiologists for the first of these stages; i.e., for the process of extracting the tangents to piecewise smooth curves from a two-dimensional image. We present an analysis of the orientation selection process from a computational perspective that is strongly influenced by various biological constraints. As such, it provides both a solid foundation for curve detection algorithms within computer vision systems and illustrates the insights that can be gained by analyzing biological vision systems. Formal extensions to the algorithm can also be posed that provide further insight into texture and optical flow.


Receptive Field Optical Flow Simple Cell Lateral Maximum Fingerprint Image 
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.

7. References

  1. 1.
    Dobbins, A., Zucker, S.W., and Cynader, M., Endstopping in the visual cortex as a neural substrate for calculating curvature, Nature, 1987, 329, 438–441.PubMedGoogle Scholar
  2. 2.
    Hubel, D., and Wiesel, T., Functional architecture of macaque monkey visual cortex, Proceedings of the Royal Society (London) B 198, 1977, pp. 1–59.Google Scholar
  3. 3.
    Gilbert, C., and Wiesel, T., Laminar specialization and intra-cortical connections in cat primary visual cortex, in F. Schmidt, F. Worden, G. Adalman, and S. Dennis (eds.), The Organization of the Cerebral Cortex. MIT Press, Cambridge, 1981.Google Scholar
  4. 4.
    Hummel, R., and Zucker, S.W., “On the foundations of relaxation labelling processes,” IEEE Trans. PAMI 5, 1983, pp. 267–287.Google Scholar
  5. 5.
    Iverson, Lee, and Zucker, S.W., Orientation selection to optical flow: a computational perspective, Proc. IEEE Workshop on Computer Vision, Miami, 1987.Google Scholar
  6. 6.
    Link, N., and Zucker, S.W., Sensitivity to corners in Flow Patterns, Spatial Vision, 1988a, to be published.Google Scholar
  7. 7.
    Link, N., and Zucker, S.W., Corner detection in curvilinear dot grouping, Biological Cybernetics, 1988b, to be published.Google Scholar
  8. 9.
    Orban, G.A., Neuronal Operations in the Visual Cortex, Springer-Verlag, 1984.Google Scholar
  9. 10.
    Parent, P. and Zucker, S.W., Radial Projection: An Efficient Update Rule for Relaxation Labelling, CVaRL Technical Report TR-86-15R, McGill University, 1985a; IEEE Trans. PAMI, in press.Google Scholar
  10. 11.
    Parent, P. and Zucker, S.W., Trace Inference, Curvature Consistency, and Curve Detection, CVaRL Technical Report CIM-86-3, McGill University, 1985b;IEEE Trans. PAMI, in press.Google Scholar
  11. 12.
    Zucker, S.W., Early Orientation Selection, Tangent Fields and the Dimensionality of their Support, Computer Vision, Graphics and Image Processing 32, 1985 pp. 74–103.Google Scholar
  12. 13.
    Zucker, S.W., The computational connection in vision: Early orientation selection. Behaviour Research Methods, Instruments, and Computers, 1986, 18, 608–617.Google Scholar
  13. 14.
    Zucker, S.W., Early vision, in S. Shapiro (ed.), The Encyclopedia of Artificial Intelligence, John Wiley, 1987.Google Scholar
  14. 15.
    Zucker, S.W. and Iverson, L., “From Orientation Selection to Optical Flow,” Computer Vision, Graphics and Image Processing, 37, January 1987.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1988

Authors and Affiliations

  • Steven W. Zucker
    • 1
    • 2
  1. 1.Computer Vision and Robotics Laboratory McGill Research Centre for Intelligent MachinesMcGill UniversityMontréalCanada
  2. 2.The Canadian Institute for Advanced ResearchCanada

Personalised recommendations