中文摘要：

图像修复是近年来图像视觉研究当中的一个热点.Criminisi算法是一种比较常用的方法.为了消除原算法当中置信度和数据项相互影响的问题,并且考虑到平滑项对图像的锐化作用,对优先权的计算重新进行了调整.而将粒子群算法运用到最佳匹配块的搜索过程当中,避免了全局搜索带来的大工作量和不准确性,提高了算法的修复效率和准确性.经过仿真实验证明,改进后的算法不仅在PSNR值上有所提高,修复效果也更符合人们的视觉需求.

英文摘要：

In recent years, image inpainting is one of the hot spots in the image vision research. Criminisi algorithm is a widely used algorithm. In order to eliminate the mutual influence during the confidence level and date terms, and considering the sharpening effect of smoothing item on images, we restructure the calculation of the priority. The particle swarm optimization algorithm is applied to the search process of the best matching patch, which avoids the heavy workload and inaccuracy caused by the global search, and improves the efficiency and accuracy of the algorithm. The simulation results show that the improved algorithm have better general performance in image completion. Not only the PSNR values are improved, but also the restoration results can better satisfy people＇s visual requirements.