破损区域分块划分的图像修复

目的：提出一个算法，使计算机能够自动修复破损区域较大且结构信息较复杂的图像。方法：本研究通过模仿手工修复破损区域较大且结构信息较复杂的图像的方法，按如下2个步骤来修复图像：1) 破损区域的划分，首先，对各断裂边界线进行匹配配对。然后，将已配对的各断裂边界线进行直接连接，从而在破损区域内形成各个待修复块。2) 各块的修复，首先，采用BSCB算法中的传输方程和扩散方程将已选邻域信息迭代传输和扩散到各块破损区域，以修复完优先级最大的各个块。然后，判断是否有次优先级的待修复块，若有，则采用边界线删除算法删除部分冗余边界线，接着按相同方法修复次优先级的待修复块；若无，则修复完成。结果：基于以上图像修复步骤，提出了破损区域分块划分的图像修复算法。将提出的该算法和其它3个算法用于修复破损区域较大且结构信息较复杂的图像，其结果显示，该算法所修复图像的PSNR值平均提高1.49db，同时，所修复图像具有较好的视觉效果。结论：和其它3个算法相比，提出的破损区域分块划分的图像修复算法更适合于修复破损区域较大且结构信息较复杂的图像。

Image inpainting algorithm based on partition block of damaged region

ObjectivesTo propose a image inpainting algorithm that can enable the computer to repair the image with a larger damaged area and more complex structure information.MethodsThis study repair the image with a larger damaged area and more complex structure information by imitating manual repair such damaged image method, and it has two steps: the division of damaged area and the repair of each block. 1) In the division process of damaged area: Firstly, the matching degree of every two damaged boundaryies are calculated and the boundaryies having greatest matching degree are made into matching pairs.Then, the well matched boundaries are directly connected, dividing the damaged region into different blocks. 2) In the repair process of each block, firstly, the transmission equation and diffusion equation of the BSCB algorithm are used to repair each block with highest priority. Secondly, our algorithm judges whether there is a block with second priority to be repaired, if yes, the boundary deletion algorithm is used to delete the redundant boundary lines, and the same method is used to repair each block; if not, the repair process is completed.ResultsBased on the above image inpainting steps, a image inpainting algorithm based on partition block of damaged region is proposed. The experimental results show that the proposed method can increase the PSNR value 1.49db and achieve better visual effect in repairing the image with a larger damaged area and more complex structure information compared with the other 3 image inpainting algorithm.ConclusionsThe proposed algorithm is more suitable for repairing the damaged image with a larger damaged area and more complex structure information compared with the other 3 image inpainting algorithm.