Figure-ground discrimination by mean field annealing

  • Laurent Hérault
  • Radu Horaud
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 588)


We formulate the figure-ground discrimination problem as a combinatorial optimization problem. We suggest a cost function that makes explicit a definition of shape based on interactions between image edges. These interactions have some mathematical analogy with interacting spin systems — a model that is well suited for solving combinatorial optimization problems. We devise a mean field annealing method for finding the global minimum of such a spin system and the method successfully solves for the figure-ground problem.


Combinatorial Optimization Problem Synthetic Image Image Element Field Annealing Recursive Neural Network 
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. 1.
    J. Canny. A Computational Approach to Edge Detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-8(6):679–698, November 1986.Google Scholar
  2. 2.
    R. Deriche. Using Canny's criteria to derive a recursively implemented optimal edge detector. International Journal of Computer Vision, 1(2):167–187, 1987.Google Scholar
  3. 3.
    L. Hérault and R. Horaud. Figure-ground discrimination: a combinatorial optimization approach. Technical Report RT 73, LIFIA-IMAG, October 1991.Google Scholar
  4. 4.
    L. Hérault and J.J. Niez. Neural Networks and Graph K-Partitioning. Complex Systems, 3(6):531–576, December 1989.Google Scholar
  5. 5.
    L. Hérault and J.J. Niez. Neural Networks and Combinatorial Optimisation: A Study of NP-Complete Graph Problems. In E. Gelembe, editor, Neural Networks: Advances and Applications, pages 165–213. North Holland, 1991.Google Scholar
  6. 6.
    W. Kohler. Gestalt Psychology. Meridian, New-York, 1980.Google Scholar
  7. 7.
    H. Orland. Mean field theory for optimization problems. Journal Physique Lettres, 46:L–763–L–770, 1985.Google Scholar
  8. 8.
    P. Parent and S.W. Zucker. Trace Inference, Curvature Consistency, and Curve Detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 11(8):823–839, August 1989.Google Scholar
  9. 9.
    C. Peterson. A new method for mapping optimization problems onto neural networks. International Journal of Neural Systems, 1(1):3–22, 1989.Google Scholar
  10. 10.
    T. J. Sejnowski and G. E. Hinton. Separating Figure from Ground with a Boltzmann Machine. In Michael Arbib and Allen Hanson, editors, Vision, Brain, and Cooperative Computation, pages 703–724. MIT Press, 1988.Google Scholar
  11. 11.
    A. Sha'ashua and S. Ullman. Structural Saliency: The Detection of Globally Salient Structures Using a Locally Connected Network. In Proc. IEEE International Conference on Computer Vision, pages 321–327, Tampa, Florida, USA, December 1988.Google Scholar
  12. 12.
    H.E. Stanley. Introduction to Phases Transitions and Critical Phenomena. Oxford University Press, 1971.Google Scholar
  13. 13.
    D.E. Van den Bout and T. K. Miller. Graph Partitioning Using Annealed Neural Networks. In Int. Joint Conf. on Neural Networks, pages 521–528, Washington D.C., June 1989.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1992

Authors and Affiliations

  • Laurent Hérault
    • 1
  • Radu Horaud
    • 2
  1. 1.CEA-LETIGrenoble
  2. 2.LIFIA-IRIMAGGrenobleFrance

Personalised recommendations