Investigating Method Of Texture Synthesis Based On Markov Random Field (MRF)

Texture synthesis has been one of the hottest areas in Computer Graphics, Computer Vision and Image Processing fields, and it has application in image repair, texture replacement, cartoon facture, image coding, and error concealment. Along with the farther research of texture analysis & synthesis, there are more and more application fields.This paper introduces firstly some notions, techniques, methods and recent development in the texture synthesis field, and explains the structure model & statistic model on texture synthesis for the sake of the following text. Studying and researching deeply the synthesis methods based on image pixels and blocks, and under Markov random field (MRF) model hypothesis the paper presents the causal neighbour field synthesis method on based image pixels, and under the same hypothesis it proposes the causal neighbour field synthesis method on based image blocks. Researching, experimenting and comparing with the new synthesis algorithm, the new algorithm has great advantage.In analyzing texture synthesis techniques based on image pixels, the paper compares the histogram matching synthesis method with the proposed causal neighbour field synthesis method, and the results show that the proposed approach vastly excels the histogram matching synthesis one special in structure texture images. In analyzing texture synthesis techniques based on image blocks, the paper compares the proposed causal neighbour field synthesis method with Andrew Nealen's method, and the results show that the proposed algorithm costs less time than Andrew Nealen's method under the same image quality.Through the research of developing and compiling the application programs of texture synthesis, experimenting image processing emulation with the Matlab software. And we apply texture synthesis techniques to texture replacement, image repair and error concealment. , and the results show the outcome is very good.In the future, texture synthesis must be applied in more and more fields. The future works about the technique of texture synthesis are introduced. We must more hard work, and step by step develop and research.