Decision Rules for Choice of Neighbors in Random Field Models of Images

R. Chellappa, R.L. Kashyap and Narendra Ahuja

Abstract

Random field models have many applications in image processing and analysis. The authors design a decision rule for fitting an appropriate random field model to a given image. They assume that the given image is a particular realization of a homogeneous Gaussian discrete random field. They represent the underlying random field by a set of parametric models representing the spatial dependence. Using spectral representations of the random field and standard Bayesian methods, the authors develop a decision rule for choosing an appropriate model from a class of such models. They discuss the relevance of the theory developed for applications in image modeling and texture characterization.