Low Complexity Image/Video Super Resolution Using Edge and Nonlocal Self-Similarity Constraint

Zongliang GAN  

Publication
IEICE TRANSACTIONS on Information and Systems   Vol.E96-D   No.7   pp.1569-1572
Publication Date: 2013/07/01
Online ISSN: 1745-1361
DOI: 10.1587/transinf.E96.D.1569
Print ISSN: 0916-8532
Type of Manuscript: LETTER
Category: Image Processing and Video Processing
Keyword: 
bilateral filtering,  self-similarity,  nonlocal means,  back projection,  super resolution,  

Full Text: PDF>>
Buy this Article




Summary: 
In this letter, we present a fast image/video super resolution framework using edge and nonlocal constraint. The proposed method has three steps. First, we improve the initial estimation using content-adaptive bilateral filtering to strengthen edge. Second, the high resolution image is estimated by using classical back projection method. Third, we use joint content-adaptive nonlocal means filtering to get the final result, and self-similarity structures are obtained by the low resolution image. Furthermore, content-adaptive filtering and fast self-similarity search strategy can effectively reduce computation complexity. The experimental results show the proposed method has good performance with low complexity and can be used for real-time environment.