An Improved algorithms for Imge inpainting
1、Department of Information and Engineering of Wuhan University of Technology
Abstract:Digital image inpainting is one of the important research fields in the computer graphics, computer vision and image procession. The primary methods of image inpainting include texture synthesis algorithms and the algorithms based on PDE. The texture synthesis based image inpainting algorithm, propose by Criminisi and his partner, is deeply researched, in view of searching the source patch globally may lead to error match, and such error accumulates and propagates to other areas quickly, even makes the results completely unacceptable, this paper uses the texture segment result of inpainting image to guide the match patch’s search, avoid to generate many errors. The experiment results show the advantages of the improved algorithm.