Image super-resolution via sparse representation
- PMID: 20483687
- DOI: 10.1109/TIP.2010.2050625
Image super-resolution via sparse representation
Abstract
This paper presents a new approach to single-image super-resolution, based on sparse signal representation. Research on image statistics suggests that image patches can be well-represented as a sparse linear combination of elements from an appropriately chosen over-complete dictionary. Inspired by this observation, we seek a sparse representation for each patch of the low-resolution input, and then use the coefficients of this representation to generate the high-resolution output. Theoretical results from compressed sensing suggest that under mild conditions, the sparse representation can be correctly recovered from the downsampled signals. By jointly training two dictionaries for the low- and high-resolution image patches, we can enforce the similarity of sparse representations between the low resolution and high resolution image patch pair with respect to their own dictionaries. Therefore, the sparse representation of a low resolution image patch can be applied with the high resolution image patch dictionary to generate a high resolution image patch. The learned dictionary pair is a more compact representation of the patch pairs, compared to previous approaches, which simply sample a large amount of image patch pairs, reducing the computational cost substantially. The effectiveness of such a sparsity prior is demonstrated for both general image super-resolution and the special case of face hallucination. In both cases, our algorithm generates high-resolution images that are competitive or even superior in quality to images produced by other similar SR methods. In addition, the local sparse modeling of our approach is naturally robust to noise, and therefore the proposed algorithm can handle super-resolution with noisy inputs in a more unified framework.
Similar articles
-
Super-resolution CT Image Reconstruction Based on Dictionary Learning and Sparse Representation.Sci Rep. 2018 Jun 11;8(1):8799. doi: 10.1038/s41598-018-27261-z. Sci Rep. 2018. PMID: 29892023 Free PMC article.
-
Image transformation based on learning dictionaries across image spaces.IEEE Trans Pattern Anal Mach Intell. 2013 Feb;35(2):367-80. doi: 10.1109/TPAMI.2012.95. IEEE Trans Pattern Anal Mach Intell. 2013. PMID: 22529324
-
Image Super-Resolution Based on Structure-Modulated Sparse Representation.IEEE Trans Image Process. 2015 Sep;24(9):2797-810. doi: 10.1109/TIP.2015.2431435. IEEE Trans Image Process. 2015. PMID: 25966473
-
Sparse signal representation and its applications in ultrasonic NDE.Ultrasonics. 2012 Mar;52(3):351-63. doi: 10.1016/j.ultras.2011.10.001. Epub 2011 Oct 10. Ultrasonics. 2012. PMID: 22040650 Review.
-
Application of Sparse Representation in Bioinformatics.Front Genet. 2021 Dec 15;12:810875. doi: 10.3389/fgene.2021.810875. eCollection 2021. Front Genet. 2021. PMID: 34976030 Free PMC article. Review.
Cited by
-
Restoration and Enhancement of Historical Stereo Photos.J Imaging. 2021 Jun 24;7(7):103. doi: 10.3390/jimaging7070103. J Imaging. 2021. PMID: 39080891 Free PMC article.
-
Super Resolution of Magnetic Resonance Images.J Imaging. 2021 Jun 21;7(6):101. doi: 10.3390/jimaging7060101. J Imaging. 2021. PMID: 39080889 Free PMC article.
-
Single Image Super-Resolution via Wide-Activation Feature Distillation Network.Sensors (Basel). 2024 Jul 16;24(14):4597. doi: 10.3390/s24144597. Sensors (Basel). 2024. PMID: 39065995 Free PMC article.
-
Single-sample image-fusion upsampling of fluorescence lifetime images.Sci Adv. 2024 May 24;10(21):eadn0139. doi: 10.1126/sciadv.adn0139. Epub 2024 May 23. Sci Adv. 2024. PMID: 38781345 Free PMC article.
-
Uncertainty-driven mixture convolution and transformer network for remote sensing image super-resolution.Sci Rep. 2024 Apr 24;14(1):9435. doi: 10.1038/s41598-024-59384-x. Sci Rep. 2024. PMID: 38658596 Free PMC article.
Publication types
LinkOut - more resources
Full Text Sources
Other Literature Sources
Research Materials