基于结构性信息的纹理合成与图像修复  

Texture synthesis and image inpainting based on a new intensity-structure difference measure.

作　　者：沈强[1] 李伟青[1] 

机构地区：[1]浙江大学CAD&CG国家重点实验室,浙江杭州310058

出　　处：《浙江大学学报（理学版）》2006年第5期510-513,共4页Journal of Zhejiang University（Science Edition）

摘　　要：纹理合成与图像修复是计算机图形学中的一个研究热点,具有广泛的应用前景.提出了一种基于结构性信息的判断纹理块匹配误差的新方法:颜色-结构差异算法.该方法同时考虑了纹理块之间颜色误差和结构误差对纹理块的相似性的影响,弥补了单纯考虑颜色误差的不足,保证了图像间结构信息的连贯性,从而确保合成后的纹理及修复后的图像在视觉上的连贯性.Texture synthesis and image inpainting are the hot research points of computer graphics and have widely application. A new difference measure for texture synthesis, intensity-structure difference （ISD）, is presented. This measure considers the intensity difference and structure difference simultaneously when computing the similarity within patches. It successfully overcomes the drawbacks the general intensity difference method has, ensuring the structure continuity within the textures and retaining perceptual coherence in synthesized texture and inpainted image.

关 键 词：马可夫随机域 纹理合成 图像修复 颜色结构差异 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]