Single image super-resolution reconstruction algorithm based on image self-similarity and multi-task Gaussian process regression

Jian-hong LI; Ya-rong WU; Ju-jian LÜ

doi:10.3788/OPE.20182611.2814

您当前的位置：

首页 >

文章列表页 >

Single image super-resolution reconstruction algorithm based on image self-similarity and multi-task Gaussian process regression

Information Sciences | 更新时间：2020-07-05

- Single image super-resolution reconstruction algorithm based on image self-similarity and multi-task Gaussian process regression
- Optics and Precision Engineering Vol. 26, Issue 11, Pages: 2814-2826(2018)
- 作者机构：
  
  1.广东外语言外贸大学语言工程与计算实验, 广东广州 510006
  2.仲恺农业工程学院仲恺科技服务公司, 广东广州 510225
  3.广东技术师范学院计算机学院, 广东广州 510665
- 作者简介：
- 基金信息：
- DOI：10.3788/OPE.20182611.2814
  CLC： TP391.4
- Received：30 May 2018，
  
  Accepted：25 June 2018，
  
  Published：25 November 2018
- 稿件说明：
移动端阅览
Jian-hong LI, Ya-rong WU, Ju-jian LÜ. Single image super-resolution reconstruction algorithm based on image self-similarity and multi-task Gaussian process regression[J]. Optics and precision engineering, 2018, 26(11): 2814-2826.
DOI：

Jian-hong LI, Ya-rong WU, Ju-jian LÜ. Single image super-resolution reconstruction algorithm based on image self-similarity and multi-task Gaussian process regression[J]. Optics and precision engineering, 2018, 26(11): 2814-2826. DOI： 10.3788/OPE.20182611.2814.

摘要

在单帧图像超分辨率问题中，基于高斯过程回归的超分辨率算法没有挖掘相似图像片间的关联关系或者无差别地用相似图像片来扩充训练集合，都会导致重建的高分辨率图像中存在明显的噪声和伪影。对此提出了一种基于多任务高斯过程回归的超分辨率算法。该算法通过引入多任务学习思想，将输入的低分辨率图像进行分片处理，把每一个图像片的超分辨率过程视为一个任务。在对相似任务建模的过程中，通过最优化求解的参数集合来体现任务间的共性及差异，从而使模型的泛化能力和预测精度得以提高，在重建高分辨率图像清晰锐利的同时，噪声和伪影受到明显抑制。用常见的测试图像以及公开的图像测试集合进行的大量试验表明该算法在主观评价和客观评价两个方面均优于同类型算法及当前经典算法，峰值信噪比较其它常见超分辨率算法可提高约0.5 dB。

Abstract

In the domain of single image super-resolution

algorithms based on Gaussian process regression neither exploit the association relationships among similar patches

nor do they discriminate between these patches with similar properties to augment the volume of the training set

which leads to obvious noise and artifacts in reconstructed high-resolution images. To overcome this problem

a new super-resolution algorithm based on multi-task Gaussian process regression is proposed. This algorithm introduces the idea of multi-task learning to partition the input low-resolution image into overlapped patches and considers the super-resolution process of each patch as a task. In the process of modeling similar tasks

the parameter set obtained by optimal solving for representing the commonness and difference gives generalization ability

improves prediction accuracy improved

makes the reconstructed high-resolution image clear and sharp

and suppresses noise and artifacts significantly. A large number of experiments to process common testing images and a public image test set subjectively and objectively demonstrate that this algorithm is superior to similar state of the art algorithms

and the peak signal to noise ratio is approximately 0.5 dB higher than that of other common super-resolution algorithms.

关键词

Keywords

references

LI X, ORCHARD M T. New edge-directed interpolation[J]. 2001, 10(10): 1521-1527. http://med.wanfangdata.com.cn/Paper/Detail/PeriodicalPaper_PM18255495

符冉迪, 周颖, 颜文, 等.基于TV-L1分解的红外云图超分辨率算法[J].光学精密工程, 2016, 24(4):937-944.

FU R D, ZHOU Y, YAN W, et al .. Infrared nephogram super-resolution algorithm based on TV-L 1 decomposition[J]. Opt. Precision Eng. , 2016, 24(4):937-944. (in Chinese)

CHAN T, ESEDOGLU S, PARK F, et al .. Total Variation Image Restoration: Overview and Recent Developments [M]. Handbook of Mathematical Models in Computer Vision, 2006:17-31.

SHAN Q, LI Z, JIA J, et al .. Fast image/video up-sampling[C]. ACM SIGGRAPH Asia, 2008: 153. https://www.researchgate.net/publication/220184280_Fast_ImageVideo_Upsampling

TAI Y W, LIU S, BROWN M S, et al .. Super resolution using edge prior and single image detail synthesis[C]. Computer Vision and Pattern Recognition , 2010: 2400-2407. https://www.researchgate.net/publication/224164286_Super_Resolution_using_Edge_Prior_and_Single_Image_Detail_Synthesis

YANG J, WANG Z, LIN Z, et al .. Coupled dictionary training for image super-resolution[J]. IEEE Transactions on Image Processing , 2012, 21(8):3467-3478.

YANG J, WRIGHT J, HUANG T S, et al .. Image super-resolution via sparse representation[J]. IEEE Transactions on Image Processing , 2010, 19(11):2861-2873.

韩玉兰, 赵永平, 王启松, 等.稀疏表示下的噪声图像超分辨率重构[J].光学精密工程, 2017, 25(6):1619-1626.

HAN Y, ZHAO Y, WANG Q, et al .. Reconstruction of super resolution for noise image under the sparse representation[J]. Opt. Precision Eng. , 2017, 25(6):1619-1626. (in Chinese)

DONG W, ZHANG L, SHI G, et al .. Image deblurring and super-resolution by adaptive sparse domain selection and adaptive regularization[J]. IEEE Transactions on Image Processing , 2011, 20(7):1838.

KIM J, LEE J K, LEE K M. Deeply-recursive convolutional network for image super-resolution[C]. IEEE Conference on Computer Vision and Pattern Recognition , 2016: 1637-1645. http://www.oalib.com/paper/4055025

KIM J, LEE J K, LEE K M. Accurate image super-resolution using very deep convolutional networks[C]. IEEE Conference on Computer Vision and Pattern Recognition , 2016: 1646-1654. http://www.oalib.com/paper/4055028

DONG C, CHEN C L, HE K, et al .. Learning a deep convolutional network for image super-resolution[C]. European Conference on Computer Vision , 2014: 184-199. https://rd.springer.com/chapter/10.1007/978-3-319-10593-2_13

WANG Z, LIU D, YANG J, et al .. Deep networks for image super-resolution with sparse prior[C]. IEEE International Conference on Computer Vision , 2015: 370-378. http://www.oalib.com/paper/4049782#.XCApUvk6vWY

GU S, ZUO W, XIE Q, et al .. Convolutional sparse coding for image super-resolution[C]. IEEE International Conference on Computer Vision. , 2015: 1823-1831. https://www.researchgate.net/publication/300412580_Convolutional_Sparse_Coding_for_Image_Super-Resolution

GLASNER D, BAGON S, IRANI M. Super-resolution from a single image[C]. International Conference on Computer Vision , 2009: 349-356. https://www.researchgate.net/publication/224135979_Super-Resolution_from_a_Single_Image?ev=auth_pub

YANG J, LIN Z, COHEN S. Fast image super-resolution based on in-place example regression[C]. Computer Vision and Pattern Recognition , 2013: 1059-1066. https://www.researchgate.net/publication/261259210_Fast_Image_Super-Resolution_Based_on_In-Place_Example_Regression

HE H, SIU W C. Single image super-resolution using Gaussian process regression[C]. IEEE Conference on Computer Vision and Pattern Recognition , 2011: 449-456. https://www.researchgate.net/publication/221363736_Single_image_super-resolution_using_Gaussian_process_regression

LI J, QU Y, LI C, et al .. Learning local Gaussian process regression for image super-resolution[J]. Neurocomputing, 2015, 154(C):284-295.

WANG H, GAO X, ZHANG K, et al .. Image super-resolution using non-local Gaussian process regression[J]. Neurocomputing , 2016, 194:95-106.

KWON Y, KIM K I, TOMPKIN J, et al .. Efficient learning of image super-resolution and compression artifact removal with semi-local Gaussian processes[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence , 2015, 37(9):1.

袁其平, 林海杰, 陈志宏, 等.用支持向量回归法实现单帧图像超分辨率重建[J].光学精密工程, 2016, 24(09):2302-2309.

YUAN Q P, LIN H J, CHEN ZH H, et al .. Single image super-resolution reconstruction using support vector regression[J]. Opt. Precision Eng. , 2016, 24(9):2302-2309. (in Chinese)

WANG S, ZHANG L, URTASUN R. Transductive Gaussian processes for image denoising[C]. IEEE International Conference on Computational Photography , 2014: 1-8. https://www.researchgate.net/publication/269302160_Transductive_Gaussian_processes_for_image_denoising

CHENG K W, CHEN Y T, Fang W H. Video anomaly detection and localization using hierarchical feature representation and Gaussian process regression[C]. IEEE Conference on Computer Vision and Pattern Recognition , 2015: 2909-2917. https://www.researchgate.net/publication/282615597_Video_Anomaly_Detection_and_Localization_Using_Hierarchical_Feature_Representation_and_Gaussian_Process_Regression

LONG C, HUA G. Multi-class multi-annotator active learning with robust Gaussian process for visual recognition[C]. IEEE International Conference on Computer Vision , 2015: 2839-2847. https://www.researchgate.net/publication/304409747_Multi-class_Multi-annotator_Active_Learning_with_Robust_Gaussian_Process_for_Visual_Recognition

RASMUSSEN C E, Williams C K I. Gaussian processes for machine learning (adaptive computation and machine learning) [M]. The MIT Press, 2005.

ZHANG Y, YEUNG D Y. Multi-task warped Gaussian process for personalized age estimation[C]. IEEE Conference on Computer Vision and Pattern Recognition , 2010: 2622-2629. https://www.researchgate.net/publication/221362463_Multi-Task_Warped_Gaussian_Process_for_Personalized_Age_Estimation

SNELSON E, RASMUSSEN C E, GHAHRAMANI Z. Warped Gaussian processes[C]. International Conference on Neural Information Processing Systems , 2004: 2003. https://www.semanticscholar.org/paper/Warped-Gaussian-Processes-Snelson-Rasmussen/13292517cd09abacba4a3ab42a79cb97c2e7526d

MAIRAL J, BACH F, PONCE J, et al .. Non-local sparse models for image restoration[C]. IEEE International Conference on Computer Vision , 2009: 2272-2279. https://www.researchgate.net/publication/221110774_Non-local_sparse_models_for_image_restoration

BUADES A, COLL B, MOREL J M. A non-local algorithm for image denoising[C]. Computer Vision and Pattern Recognition , 2005: 60-65. https://www.researchgate.net/publication/4156453_A_non-local_algorithm_for_image_denoising

LAWRENCE N, SEEGER M, HERBRICH R. Fast sparse Gaussian process methods: the informative vector machine[C]. International Conference on Neural Information Processing Systems , 2002: 625-632. https://www.researchgate.net/publication/221618434_Fast_Sparse_Gaussian_Process_Methods_The_Informative_Vector_Machine

HUANG J B, SINGH A, AHUJA N. Single image super-resolution from transformed self-exemplars[C]. IEEE Conference on Computer Vision and Pattern Recognition , 2015: 5197-5206. Single image super-resolution from transformed self-exemplars

TIMOFTE R, SMET V D, GOOL L V. A+: adjusted anchored neighborhood regression for fast super-resolution[C]. Asian Conference on Computer Vision , 2014: 111-126. https://link.springer.com/chapter/10.1007/978-3-319-16817-3_8

Views

193

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Radiation measurement of small targets based on PSF

Related Author

CHANG Song-tao

SUN Zhi-yuan

ZHANG Yao-yu

ZHU Wei

Related Institution

Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences

University of Chinese Academy of Sciences

AI问答

Address：No.3888 Dong Nanhu Road, Changchun, Jilin, China Postal code：130033
Tel：0431-86176855 Email：gxjmgc@ciomp.ac.cn
Technical support is provided by Beijing Founder electronics co., LTD 吉ICP备11002662号-17 京公网安备11010802024621
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰