Abstract
The update equations for Kittler Hancock probabilistic relaxation are solved exactly for a model problem. The solution reveals deficiencies of the support function. We formulate a new form of probabilistic relaxation heuristically based on the Baum Eagon theorem. The new algorithm gives the MAP labelling for the model problem.
Preview
Unable to display preview. Download preview PDF.
References
R. A. Hummel, S. W. Zucker, On the foundations of Relaxation Labeling Processes, IEEE Pattern Analysis and Machine Intelligence 5(3), 267–287, (1983).
J. Kittler, E. R. Hancock, Combining evidence in probabilistic relaxtion, Int. J. of Pattern Recognition and Artificial Intelligence 3(1), 25–51, (1989).
J. Kittler, J. Foglein, On compatibility and support functions in probabilistic relaxation, Computer Vision, Graphics and Image Processing, 34, 257–267, (1986).
R. Szelski, Bayesian modeling of uncertainty in low-level vision, Kluwer, (1989).
L. E. Baum, J. A. Eagon, An inequality with applications to statistical estimation ..., Bulletin of the American Mathematical Society, 73 360–363 (1967).
M. Pelillo, On the dynamics of relaxation labelling processes, ICNN (1994).
An extended version of this paper can be found at ftp.ee.surrey.ac.uk in directory /pub/vision/papers.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1995 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Stoddart, A.J., Petrou, M., Kittler, J. (1995). A new algorithm for probabilistic relaxation based on the Baum Eagon theorem. In: Hlaváč, V., Šára, R. (eds) Computer Analysis of Images and Patterns. CAIP 1995. Lecture Notes in Computer Science, vol 970. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-60268-2_363
Download citation
DOI: https://doi.org/10.1007/3-540-60268-2_363
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-60268-2
Online ISBN: 978-3-540-44781-8
eBook Packages: Springer Book Archive