AN EMPIRICAL-STUDY OF THE SIMULATION OF VARIOUS MODELS USED FOR IMAGES

Markov random fields are typically used as priors in Bayesian image re storation methods to represent spatial information in the image. Commo nly used Markov random fields are not in fact capable of representing the moderate-to-large scale clustering present in naturally occurring images and can also be time consuming to simulate, requiring iterative algorithms which can take hundreds of thousands of sweeps of the imag e to converge. Markov mesh models, a causal subclass of Markov random fields, are, however, readily simulated. We describe an empirical stud y of simulated realizations from various models used in the literature , and we introduce some new mesh-type models. We conclude, however, th at while large-scale clustering may be represented by such models, str ong directional effects are also present for all but very limited para meterizations. It is emphasized that the results do not detract from t he use of Markov random fields as representers of local spatial proper ties, which is their main purpose in the implementation of Bayesian st atistical approaches to image analysis. Brief allusion is made to the issue of parameter estimation.