Abstract

Markov random fields are used extensively in modelbased approaches to image segmentation and, under the Bayesian paradigm, are implemented through Markov chain Monte Carlo (MCMC) methods. In this paper, we describe a class of such models (the double Markov random field) for images composed of several textures, which we consider to be the natural hierarchical model for such a task. We show how several of the Bayesian approaches in the literature can be viewed as modifications of this model, made in order to make MCMC implementation possible. From a simulation study, conclusions are made concerning the performance of these modified models.

Citations

3012	Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images - Geman, Geman - 1984
910	Spatial interaction and the statistical analysis of lattice systems - Besag - 1974
573	A maximization technique occurring in the statistical analysis of probabilistic functions of Markov chains - Baum, Petrie, et al. - 1970
233	Exact maximum a posteriori estimation for binary images - Greig, Porteous, et al. - 1989
203	Nonuniversal critical dynamics in Monte Carlo simulations - Swendsen, Wang - 1987
167	Monte Carlo Maximum Likelihood for Depend Data - Geyer
141	CJ: Markov chain Monte Carlo maximum likelihood - Geyer
56	Estimation and choice of neighbors in spatialinteraction models of images - KASHYAP, R - 1984
50	simple parallel hierarchical and relaxation algorithms for sementing non-causal Markovian random fields - Cohen, Cooper - 1987
50	Fast sampling of Gaussian Markov random fields - Rue - 2001
42	On block updating in Markov random field models for disease mapping - Knorr-Held, Rue - 2002
33	Segmentation of textured images using a multiresolution Gaussian autoregressive model - Comer, Delp - 1999
27	Stochastic and deterministic networks for texture segmentation - Manjunath, Simchony, et al. - 1990
25	Segmentation of textured images using Gibbs random fields - Derin, Cole - 1986
17	Spatial applications of Markovchain Monte Carlo for Bayesian inference - Higdon - 1994
15	An empirical study of the simulation of various models used for images - Gray - 1994
15	Fully Bayesian approach to image restoration with an application in biogeography - Heikkinen, Hogmander - 1994
14	Image Segmentation using Markov Random Field Models - Barker - 1998
13	New loss functions in Bayesian imaging - Rue - 1995
10	Fully bayesian reconstructions from single-photon emission computed tomography data - Weir - 1997
7	Valid parameter space for 2-D Gaussian Markov random fields - Lakshmanan, Derin - 1993
6	Bayesian color image segmentation using reversible jump Markov chain Monte Carlo, Research Report 01/99-R055 - Kato - 1999
3	Graffigne “Markov Random Field Models and Their Applications to Computer Vision - Geman, C - 1987
2	Dynamic image analysis using Bayesian shape and texture models - Phillips, Smith - 1993
2	Bayesian image classification using Baddeley’s delta loss - Rue, Frigessi - 1997
1	Maximum likelihood estimation for unsupervised model-based image segmentation - Zhang, Modestino, et al. - 1994
1	Modeling and segmentation of noisy and textured images using random fields - Derin, Elliott - 1987
1	A Bayesian approach to the segmentation of textured images - Melas - 1999