Title :
Image Restoration Based on Eigensubspace for Image Structure
Author :
Hu, Xiaotong ; Cao, Dongrui ; Qiu, Shuping
Author_Institution :
Tianjin Univ. of Sci. & Technol.
Abstract :
The image restoration is the technology to restore the damaged part in the image automatically. In this paper we proposed a new restoration method based on the eigensubspace for image structure. After the regions need to be restored are selected, the proposed method use the undamaged part to make up the eigenspace, and then, search different set of image blocks that have high similarity with for each damaged image block, to make up eigensubspace for image structure. At the end, the damaged image blocks can be restored according to their eigensubspace. Experiments result indicates that the proposed method can restore damaged image accurately,
Keywords :
eigenvalues and eigenfunctions; image restoration; eigensubspace; image restoration; image structure; Fractals; Image coding; Image restoration; Pixel;
Conference_Titel :
Multimedia, 2006. ISM'06. Eighth IEEE International Symposium on
Conference_Location :
San Diego, CA
Print_ISBN :
0-7695-2746-9
DOI :
10.1109/ISM.2006.90