Bayesian Estimation of a Random Field with Discontinuities

This paper deals with the estimation problem of a random field with discontinuities. The method is based on the Bayesian approach. When the Bayesian method is applied to the estimation problem of a random field, a prior information about the random field plays a crucial role. It is revealed that in the case of the random field with discontinuities, as the prior information, the Cauchy probability distribution for the spatial variation of the random field is superior to the Gaussian probability distribution which is widely used. Then, it is shown that the estimation algorithm obtained is realized in a parallel manner using a neural network model.

[1]  A. N. Tikhonov,et al.  REGULARIZATION OF INCORRECTLY POSED PROBLEMS , 1963 .

[2]  Demetri Terzopoulos,et al.  Regularization of Inverse Visual Problems Involving Discontinuities , 1986, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  C Koch,et al.  Analog "neuronal" networks in early vision. , 1986, Proceedings of the National Academy of Sciences of the United States of America.

[4]  Jitendra Malik,et al.  Scale-Space and Edge Detection Using Anisotropic Diffusion , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[5]  Tomaso Poggio,et al.  Computational vision and regularization theory , 1985, Nature.

[6]  Donald Geman,et al.  Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.