期刊名称：International Journal of Multimedia and Ubiquitous Engineering
印刷版ISSN：1975-0080
出版年度：2015
卷号：10
期号：4
页码：193-204
DOI：10.14257/ijmue.2015.10.4.19
出版社：SERSC
摘要：Image completion is aim to fill the missing regions in images. A robust completion technique using similarity analysis and transformation is proposed to address this problem. Firstly, in order to decrease the search space of patches, random mapping method is used to analyze texture regions which have similar structure and texture with damaged regions. Secondly, geometric and photometric transformations of image are adopted to find the best patches. Thirdly, increasing the accuracy of the structure propagation, a priority calculation method is optimized based on confidence factor and edge information. Finally, a number of examples on real and synthetic images show the effectiveness of our algorithm for image completion.
