一种改进的基于纹理合成的数字图像修复方法  

An Improved Digital Image Inpainting Algorithm Based on Texture Synthesis

作　　者：桂要生[1] 邹孝[1] 肖盛[1] 戴苏秦 钱盛友[1] 

机构地区：[1]湖南师范大学物理与信息科学学院,长沙410081

出　　处：《计算机与数字工程》2015年第9期1651-1654,共4页Computer & Digital Engineering

基　　金：国家自然科学基金项目(编号:11174077;11474090);湖南省自然科学基金项目(编号:11JJ3079);湖南省教育厅基金项目(编号:12C0237;12C0204)资助

摘　　要：文章介绍了纹理合成技术中的Criminisi算法,针对其在计算边界优先权时存在的一些不足,提出了一种改进方法,通过优化Criminisi算法中优先权的计算公式,可以灵活地对不同特征的图像选择不同的参数,使得其修复结果更为合理,从而扩大了可以修复的图像范围。实验结果证明,改进方法较好地克服Criminisi算法在计算边界优先权时存在的不足,图像修复后的视觉效果和客观评价体系均较之前有所提高。The Criminisi algorithm in the technology of texture synthesis is introduced in the article, as it has some deficiencies in the calculation of the boundary priorities, an improved method is proposed, by optimizing calculation formula of the priorities in the Criminisi algorithm, it can flexibly choose different parameters according to the images with different characteristics, so as to makes the repair result more reasonable, thus expanding the scope of images which can be restored. The experimental results show that the improved method can overcome the shortages of Criminisi algorithm in the calculation of the boundary priorities, and the vis- ual effect of restored image and its objective evaluation system are improved than befor~

关 键 词：纹理合成 优先权 置信度 图像修复 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]