Image super-resolution via dual-dictionary learning and sparse representation

Learning-based image super-resolution aims to reconstruct high-frequency (HF) details from the prior model trained by a set of high- and low-resolution image patches. In this paper, HF to be estimated is considered as a combination of two components: main high-frequency (MHF) and residual high-frequ...

Full description

Saved in:
Bibliographic Details
Published in2012 IEEE International Symposium on Circuits and Systems (ISCAS) pp. 1688 - 1691
Main Authors Zhang, Jian, Zhao, Chen, Xiong, Ruiqin, Ma, Siwei, Zhao, Debin
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.05.2012
Subjects
Dictionaries
dictionary learning
image interpolation
Image reconstruction
Image resolution
Interpolation
PSNR
Signal resolution
sparse representation
super-resolution
Training
Online AccessGet full text

Cover

More Information
Summary:Learning-based image super-resolution aims to reconstruct high-frequency (HF) details from the prior model trained by a set of high- and low-resolution image patches. In this paper, HF to be estimated is considered as a combination of two components: main high-frequency (MHF) and residual high-frequency (RHF), and we propose a novel image super-resolution method via dual-dictionary learning and sparse representation, which consists of the main dictionary learning and the residual dictionary learning, to recover MHF and RHF respectively. Extensive experimental results on test images validate that by employing the proposed two-layer progressive scheme, more image details can be recovered and much better results can be achieved than the state-of-the-art algorithms in terms of both PSNR and visual perception.
ISBN:9781467302180
146730218X
ISSN:0271-4302
2158-1525
DOI:10.1109/ISCAS.2012.6271583