Perspective matching using the EM algorithm

  • A. D. J. Cross
  • E. R. Hancock
Session 6: Matching & Recognition
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1310)


This paper describes a new approach to perspective matching which simultaneously exploits both rigidity-structure and point distribution information. The structural component of the model is represented by a Delaunay triangulation of the point-set. The point-distribution model is represented by a perspective deformation of the point-set. Model-matching is realised using a variant of the EM algorithm. This involves coupling the correspondence matching of the Delaunay triangulation to the recovery of the point deformation parameters. We use a Bayesian consistency measure to gauge the relational structure of the point correspondences. Maximum-likelihood point deformation parameters are estimated using a mixture-model defined over the point error-residuals. In effect, the Bayesian consistency measure is used to weight the contributions to a mean-squares error-criterion. The method is evaluated on matching 2D objects under varying pose.


  1. 1.
    N Ahuja, “Dot Pattern Processing using Voronoi Neighbourhoods”, IEEE PAMI, 4, pp 336–343, 1982.Google Scholar
  2. 2.
    J-D. Boissonnat, “Geometric Structures for Three-Dimensional Shape Representation”, ACM Transactions on Graphics, 3, pp. 266–286, 1984.Google Scholar
  3. 3.
    T.F Cootes, C.J. Taylor, D.H. Cooper and J. Graham, “Active Shape Models-Their Training and Application”, Computer Vision, Graphics and Image Understanding, 61, pp. 38–59, 1995.Google Scholar
  4. 4.
    A.P. Dempster, Rubin N.M. and Rubin D.B., “Maximum-likelihood from incomplete data via the EM algorithm”, J. Royal Statistical Soc. Ser. B (methodological), 39, pp 1–38, 1977.Google Scholar
  5. 5.
    M.I. Jordan and R.A. Jacobs, “Hierarchical Mixtures of Experts and the EM Algorithm”, Neural Computation, 6, pp. 181–214, 1994.Google Scholar
  6. 6.
    D.P. McReynolds and D.G. Lowe, “Rigidity Checking of 3D Point Correspondences under Perspective Projection”, IEEE PAMI, 18, pp. 1174–1185, 1996.Google Scholar
  7. 7.
    G. L. Scott and H.C. Longuet-Higgins, “An Algorithm for Associating the Features of 2 Images”, Proceedings of the Royal Society of London Series B-Biological, 244, pp. 21–26, 1991.Google Scholar
  8. 8.
    S. Sclaroff and A.P. Pentland, “Modal Matching for Correspondence and Recognition”, IEEE PAMI, 17, pp. 545–661, 1995.Google Scholar
  9. 9.
    L.S. Shapiro and J. M. Brady, “Feature-based Correspondence-An Eigenvector Approach”, Image and Vision Computing, 10, pp. 283–288, 1992.Google Scholar
  10. 10.
    M. Turner and E.R. Hancock, “An EM-like Relaxation Operator”, Proceedings of the Thirteenth International Conference on Pattern Recognition, Volume II, pp. 166–170, 1996.Google Scholar
  11. 11.
    S. Ullman, “The Interpretation of Visual Motion”, MIT Press, 1979.Google Scholar
  12. 12.
    R.C. Wilson, A.N. Evans and E.R. Hancock, “Relational Matching by Discrete Relaxation”, Image and Vision Computing, 13, pp. 411–421, 1995.Google Scholar
  13. 13.
    R.C. Wilson and E.R. Hancock, “Relational Matching with Dynamic Graph Structures”, Proceedings of the Fifth International Conference on Computer Vision, pp. 450–456, 19Google Scholar
  14. 14.
    R.C. Wilson and E.R. Hancock, “Gauging Relational Consistency and Correcting Structural Errors”, IEEE Computer Society Computer Vision and Pattern Recognition Conference, pp. 54–62, 1996.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1997

Authors and Affiliations

  • A. D. J. Cross
    • 1
  • E. R. Hancock
    • 1
  1. 1.Department of Computer ScienceUniversity of YorkYorkUK

Personalised recommendations