Texture modeling and pattern analysis using statistical approach

Abstract

Texture analysis plays an important role in computer vision and pattern recog-nition. The development of a new texture model, multiresolution Markov random field (MRMRF) model, and its application in texture classification, segmentation, and textured image retrieval are the central theme of this thesis. Filtering theory and Markov random field (MRF) model are fused together seamlessly by using a new energy function. The new energy function for the texture model is constructed based on Gibbs random field.