Image completion based on texture regularity and texture synthesis
Hao C.; Yan P.; Wu E.
Source Publication5th International Conference on Information Technology and Applications, ICITA 2008
AbstractAn improved method is proposed for completing the missing parts caused by removal of foreground from an image. The research objective is to extract the foreground or background automatically and to fill the missing parts that are left behind in a very visually plausible way. In most former researches, target object is selected by users, which contains little texture information, like an apple, a person and so on. In this paper, we consider a class of target objects containing texture structures, such as a fence. Besides, filling has been addressed by two classes of algorithms in terms of texture synthesis algorithm and inpainting technique. Combining the two approaches, we integrate priority calculation with exemplar-based texture synthesis algorithms to fill in the missing regions. This combination gives each candidate patch a priority according to which synthesis algorithm will execute. Moreover, the simultaneous propagation of texture and structure information is further achieved by this priority algorithm. We demonstrate our method by completion of a variety of examples.
KeywordFilling Foreground removal Image completion Inpainting Near-regular texture Texture regularity Texture synthesis
URLView the original
Fulltext Access
Document TypeConference paper
CollectionUniversity of Macau
AffiliationUniversidade de Macau
Recommended Citation
GB/T 7714
Hao C.,Yan P.,Wu E.. Image completion based on texture regularity and texture synthesis[C],2008:53-58.
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Hao C.]'s Articles
[Yan P.]'s Articles
[Wu E.]'s Articles
Baidu academic
Similar articles in Baidu academic
[Hao C.]'s Articles
[Yan P.]'s Articles
[Wu E.]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Hao C.]'s Articles
[Yan P.]'s Articles
[Wu E.]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.