Abstract
This paper describes a new approach to extracting affine structure from 2D point-sets. The novel feature is to unify the tasks of estimating transformation geometry and identifying point-correspondence matches. Unification is realised by constructing a mixture model over the bi-partite graph representing the correspondence match and by effecting optimisation using the EM algorithm. According to our EM framework the probabilities of structural correspondence gate contributions to the expected likelihood function used to estimate maximum likelihood affine parameters. This provides a means of rejecting structural outliers. We evaluate the technique on the matching of different affine views of 3.5in floppy discs. We provide a sensitivity study based on synthetic data.
Chapter PDF
Keywords
- Delaunay Triangulation
- Point Correspondence
- Structural Match
- Affine Parameter
- Maximum Likelihood Parameter
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.
References
Y. Amit and A. Kong, “Graphical Templates for Model Registration”, IEEE PAMI, 18, pp. 225–236, 1996.
A.D.J. Cross and E.R. Hancock, “Perspective Pose Recovery with a Dual Step EM Algorithm”, Advances in Neural Information Processing Systems, 10, Edited by M. Jordan, M. Kearns and S. Solla, MIT Press, 1998.
A.P. Dempster, Laird 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.
O.D. Faugeras, E. Le Bras-Mehlman and J-D. Boissonnat, “Representing Stereo Data with the Delaunay Triangulation”, Artificial Intelligence, 44, pp. 41–87, 1990.
Gold S., Rangarajan A. and Mjolsness E., “Learning with pre-knowledge: Clustering with point and graph-matching distance measures”, Neural Computation, 8, pp. 787–804, 1996.
R.I. Hartley, “Projective Reconstruction and Invariants from Multiple Images”, IEEE PAMI, 16, pp. 1036–1041, 1994.
M.I. Jordan and R.A. Jacobs, “Hierarchical Mixtures of Experts and the EM Algorithm”, Neural Computation, 6, pp. 181–214, 1994.
M. Lades, J.C. Vorbruggen, J. Buhmann, J. Lange, C. von der Maalsburg, R.P. Wurtz and W. Konen, “Distortion-invariant object-recognition in a dynamic link architecture”, IEEE Transactions on Computers, 42, pp. 300–311, 1993
D.P. McReynolds and D.G. Lowe, “Rigidity Checking of 3D Point Correspondences under Perspective Projection”, IEEE PAMI, 18, pp. 1174–1185, 1996.
S. Moss and E.R. Hancock, “Registering Incomplete Radar Images with the EM Algorithm”, Image and Vision Computing, 15, 637–648, 1997.
D. Oberkampf, D.F. DeMenthon and L.S. Davis, “Iterative Pose Estimation using Coplanar Feature Points”, Computer Vision and Image Understanding, 63, pp. 495–511, 1996.
C. J. Poelman and T. Kanade, “A Para-perspective Factorization Method for Shape and Motion Recovery”, IEEE PAMI, 19, 1997.
P. Torr, A. Zisserman and S.J. Maybank, “Robust Detection of Degenerate Configurations for the Fundamental Matrix”, Proceedings of the Fifth International Conference on Computer Vision, pp. 1037–1042, 1995.
M. Tuceryan and T Chorzempa, “Relative Sensitivity of a Family of Closest Point Graphs in Computer Vision Applications”, Pattern Recognition, 25, pp. 361–373, 1991.
J. Utans, “Mixture Models and the EM Algorithms for Object Recognition within Compositional Hierarchies”, ICSI Berkeley Technical Report, TR-93-004, 1993.
R.C. Wilson and E.R. Hancock, “Structural Matching by Discrete Relaxation”, IEEE PAMI, 19, pp. 634–648, 1997.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1998 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Cross, A.D.J., Hancock, E.R. (1998). Holistic matching. In: Burkhardt, H., Neumann, B. (eds) Computer Vision — ECCV’98. ECCV 1998. Lecture Notes in Computer Science, vol 1407. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0054738
Download citation
DOI: https://doi.org/10.1007/BFb0054738
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-64613-6
Online ISBN: 978-3-540-69235-5
eBook Packages: Springer Book Archive