首页 > 论文 > 光学 精密工程 > 19卷 > 8期(pp:1982-1989)

运动模糊退化图像的双字典稀疏复原

Dual dictionary sparse restoration of blurred images

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

为了消除图像中的运动模糊, 提出了一种稀疏理论框架下的双字典稀疏复原方法, 并分析了冗余字典的选取和迭代算法的实现。首先, 建立了稀疏变换下的退化和复原模型, 用Haar系数冗余字典将图像稀疏化, 并用PCD阈值迭代算法对模糊图像进行收敛, 得到复原图像。由于在有效去除复原图像的模糊的同时噪声在迭代过程中被放大并叠加在图像上, 故从清晰图像库中训练了一个冗余字典进行第二次稀疏收敛来去除去模糊中被加权的噪声。实验结果表明, 本文的方法对模糊退化图像有很好的复原效果, 不仅有效地去除了运动模糊和噪声, 并能在一定程度上保留边缘细节。最后拓展了两层稀疏优化模型, 为以后在稀疏框架下的图像复原提供了新的思路。

Abstract

An image restoration method based on a dual dictionary was presented under the framework of sparse theory,and the choice of overcomplete dictionaries and the implementation of iteration methods were analyzed. Firstly, the degradation and the restoration models in the sparse theory were established,then the dictionary constructed by Haar coefficients was used to sparse the blurred image and shrink the image with Parallel Coordinate Decent(PCD) iteration algorithm to obtain the elementary deblurred image,in which the blur was removed efficiently, but the noise was weighted and added. For removing the weighted noise, the secondary dictionary from an image database was trained to shrink the deblurred image and get the final result. The results shows that the proposed method can restore the motion-blurred image efficiently, remove motion blur and noise and reserve the edge detail in some extents. Finally the two-level sparse optimization model was expanded and a new idea for the image restoration was presented under the sparse framework.

Newport宣传-MKS新实验室计划
补充资料

中图分类号:TP391.4

DOI:10.3788/ope.20111908.1982

所属栏目:信息科学

基金项目:国家自然科学基金重点项目(科学仪器专项)(No.61027002); 国家973重点基础研究发展计划资助项目(No. 2009CB72400603); 国家自然科学基金资助项目(No.60972100)

收稿日期:2011-06-09

修改稿日期:2011-07-01

网络出版日期:--

作者单位    点击查看

冯亮:北京理工大学 光电学院, 北京 100081北京理工大学 光电成像技术与系统教育部重点实验室, 北京 100081
王平:北京理工大学 光电学院, 北京 100081
许廷发:北京理工大学 光电学院, 北京 100081北京理工大学 光电成像技术与系统教育部重点实验室, 北京 100081
石明珠:北京理工大学 光电学院, 北京 100081北京理工大学 光电成像技术与系统教育部重点实验室, 北京 100081
赵峰:桂林电子科技大学, 广西 桂林 541004

联系人作者:冯亮(finalion@bit.edu.cn)

备注:冯亮(1985-), 男, 山东济南人, 博士研究生, 2007年于山东大学获得学士学位, 2009年于北京理工大学获得硕士学位, 主要从事图像复原和图像识别方面的研究。

【1】HAMAMOTO T,AIZAWA K. A computational image sensor with adaptive pixel-based integration time[J]. IEEE Journal of Solid-State Circuits, 2001, 36(4): 580-585.

【2】BEN-EZRA M,NAYAR S K. Motion-based motion deblurring[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2004, 26(6): 689.

【3】张欢. 运动模糊图像复原的全变分方法研究[D]. 西安: 西北大学, 2009.
ZHANG H. Research of total variance method for restoration of motion-blurred images[D]. Xi''an: Northwest University, 2009. (in Chinese)

【4】CALVETTI D, LEWIS B, REICHEL L. Restoration of images with spatially variant blur by the GMRES method[J]. SPIE, 2000, 4116: 364-374.

【5】YUAN L,SUN J,QUAN L, et al.. Image deblurring with blurred/noisy image pairs[C]. Proceedings of ACM SIGGRAPH,2007,26(3): 1-10.

【6】OLSHAUSEN B A. Emergence of simple-cell receptive field properties by learning a sparse code for natural images[J]. Nature, 1996,381(6583): 607-609.

【7】OLSHAUSEN B A,FIELD D J. Sparse coding with an overcomplete basis set: A strategy employed by V1[J]. Vision research, 1997, 37(23): 3311-3325.

【8】MA J,LE DIMET F X. Deblurring from highly incomplete measurements for remote sensing[J]. Geoscience and Remote Sensing, IEEE Transactions on,2009, 47(3): 792-802.

【9】ELAD M,FIGUEIREDO M A T,MA Y. On the role of sparse and redundant representations in image processing[J].IEEE, 2010,98(6): 972-982.

【10】MALLAT S G,ZHANG Z. Matching pursuits with time-frequency dictionaries[J]. IEEE Transactions on Signal Processing, 1993,41(12): 3397-3415.

【11】BERGEAUD F,MALLAT S. Matching pursuit of images [J]. Wavelet Analysis and Its Applications, 1998, 7: 285-300.

【12】PATI Y C,REZAIFAR R,KRISHNAPRASAD P. Orthogonal matching pursuit: recursive function approximation with applications to wavelet decomposition [C]. Conference on Signals,Systems and Computers,1993: 40-44.

【13】DONOHO D L,ELAD M,TEMLYAKOV V N. Stable recovery of sparse overcomplete representations in the presence of noise[J]. IEEE Transactions on Information Theory, 2006,52(1): 6-18.

【14】RANDALL P A. Sparse recovery via convex optimization [D]. California Institute of Technology, 2009.

【15】CHEN S S, DONOHO D L,SAUNDERS M A. Atomic decomposition by basis pursuit [J]. SIAM Journal on Scientific Computing, 1999, 20(1): 33-61.

【16】浦剑, 张军平. 基于词典学习和稀疏表示的超分辨率方法[J]. 模式识别与人工智能, 2010,23(3): 335-340.
PU J, ZHANG J P. Super-resolution through dictionary learning and sparse representation [J]. Pattern Recognition and Artificial Intelligence, 2010, 23(3): 335-340. (in Chinese)

【17】肖泉, 丁兴号, 王守觉, 等. 基于自适应超完备稀疏表示的图像去噪方法[J]. 仪器仪表学报, 2009, 30(9): 1886-1890.
XIAO Q, DING X H, WANG SH J, et al.. Image denoising based on adaptive over—complete sparse representation [J]. Chinese Joumal of Scientific Instrument, 2009, 30(9): 1886-1890. (in Chinese)

【18】FADILI M, STARCK J L. Sparse representation-based image deconvolution by iterative thresholding[J/OL]http: //www.oamp.fr/conf/ada4/pub/19/S/ADAIV_Fadili.pdf. 2007.

【19】ELAD M. Sparse and Redundant Representations: From Theory to Applications in Signal and Image Processing[M].New York: Springer Verlag,2010.

【20】ELAD M,AHARON M. Image denoising via sparse and redundant representations over learned dictionaries[J]. IEEE Transactions on Image Processing, 2006,15(12): 3736-3745.

【21】薛明. 压缩感知及稀疏性分解在图像复原中的应用研究[D]. 西安: 西安电子科技大学, 2009.
XUE M. Aplication research of compressive sensoring and sparse decomposition in image restoration [D]. Xi''an: Xidian University, 2009. (in Chinese)

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF