INTRODUCTION TO THE MARKOV AND GIBBS RANDOM FIELDS

Fátima Sombra de Medeiros


I. INTRODUCTION

Markov and Gibbs random fields (MRF and GRF) are currently under investigation by many research groups in image processing. The MRF and GRF approaches have attractive features in terms of statistical inference in image applications . Random fields models allow the introduction of spatial context into pixel labeling problems, such as segmentation. The models can describe textured images and lead to algorithms for generating , classifying and segmenting textured images. We are interested in how random fields can be applied to labeling or classifying images (segmentation).

II. DEFINITIONS

III. MRFs and GRFs

IV. The ICM ALGORITHM

V. BIBLIOGRAPHY

VI. ACKNOWLEDGMENTS