一种改进的基于纹理合成的图像修复算法  

An improved image inpainting algorithm based on texture synthesis

作　　者：黄淑兵[1] 朱晓临[1] 许云云[1] 朱坤[1] 

机构地区：[1]合肥工业大学数学学院,安徽合肥230009

出　　处：《合肥工业大学学报（自然科学版）》2011年第2期313-316,320,共5页Journal of Hefei University of Technology：Natural Science

基　　金：教育部科学技术研究重大资助项目(309017);安徽省自然科学基金资助项目(11040606M06)

摘　　要：文章介绍了纹理合成技术中的Criminisi算法,并对其计算边界点优先权的算法模型作了改进;针对Criminisi算法计算优先权时只考虑信任项和数据项的缺陷,增加了边界因素对优先权的影响,使得修复结果更为合理;改进的计算优先权模型灵活性较强,可以对不同特征的图像选择不同的参数,从而扩大了可以修复的图像范围。实验表明,改进的算法更为合理,修复后图像的视觉效果更好。This paper introduces the Criminisi algorithm in the texture synthesis technology,and presents an improved algorithm for computing the priority values on the boundary.The improved algorithm takes the boundary information into account when computing the priority values while the Criminisi algorithm only considers the confidence term and the data term.The results gained by the improved algorithm are much better than those of the traditional ones.And different parameters in the improved algorithm can be set to process different types of images,which makes the algorithm flexible and have a wide range of application.The examples show that the improved algorithm is reasonable and can offer effective and stable treatment of image information for damaged area.

关 键 词：纹理合成 对象移除 优先权 图像修复 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]