Two-dimensional Texture Synthesis Technology Research And Application

Texture, texture synthesis and their applications have been one of the hottest topics in the areas of Computer Graphics, Computer Vision, and Image Processing etc. Among the texture-related researches, texture synthesis techniques which address the distortion in texture mapping are most extensively used. This paper mainly focuses on the algorithm and application of two-dimensional texture synthesis. We analyses from synthesis quality and algorithm's automation, presents texture synthesis algorithm optimized by graph theory and automatic extracting texture patch size. Then we also researches several important applications of texture synthesis.Texture synthesis algorithm optimized by graph theory is an effective texture synthesis algorithm. In texture synthesis, the matching problem of overlap among textures is actually image matching problem. To resolve image matching problem, a major direction is to minimize certain important energy functions in the recent years. The method of energy functions optimized by graph cut theory is important. In this paper, we introduce min-cut/max-flow algorithm on flow of graph theory to texture synthesis domain and use it to optimize the matching of textures.Among patch-based texture synthesis algorithms, the size of the texture patch has major influence on synthesis quality. It affects how well the synthesized texture captures the local characteristics of the input sample texture. The synthesis result will not remain the features of the input texture because of either too big or too small this patch. By local and stationary of texture image, we carry out statistical analysis with a lot of texture images and find some laws, then we propose a algorithm of texel automatic extract of texture images.Textures taken from the nature world are often found inadequate to meet the requirements and we need to edit the textures to generate different textures for people's need. How to generate new texture from existing texture samples is of great significance. In this paper, we propose improved algorithms of texture synthetic application in the texture transfer and constrained single-sample texture synthesis area. The two texture synthetic application algorithms are fast and are simple to implement.