文献类型：期刊文献

英文题名：The design and implementation of GR-MV image inpainting

作者：Yuan, Jiazheng[1]; Liu, Hongzhe[1]; Zheng, Yongrong[1]

通讯作者：Yuan, J.

机构：[1] Beijing Key Laboratory of Information Service Engineering, Beijing Union University, Beijing 100101, China

第一机构：北京联合大学北京市信息服务工程重点实验室

年份：2013

卷号：9

期号：1

起止页码：72-80

外文期刊名：Journal of Computational Information Systems

收录：EI(收录号：20130515954088)

语种：英文

外文关键词：Algorithms - Gold - Image reconstruction

摘要：To solve the defects of the traditional image restoration algorithm, we proposed the golden ratio-mean variance(GR-MV) image inpainting algorithm. The algorithm is using adjacent pixels highly relevant principles. According to the golden ratio to select the local area of image, and get its mean variance, as the pixel related gradient. To analyse the information on injured pixel by the structure of the image information and texture information, thus finished the broken image inpainting. The experimental results showed that GR-MV algorithm is simple easy to understand and realized and the inpainting results are perfect. There are the characteristics which strong image inpainting and small deviations and high intelligence. Establish a theoretical foundation for the golden ratio applied to image restoration techniques, there is a great practical value. ? 2013 Binary Information Press.