Textile Texture Modeling Using Random Field Models

摘要

Described in the paper is the method to fit autoregressive(AR)random field models and Gausslan Markov ran-dom field(GMRF)models for textile textures.Maxi-mum likelihood(ML)estimation schemes and texturesynthesis for these models are presented.Three types oftypical textile textures are tested using the method andsatisfactory results are achieved.