计算机科学 ›› 2015, Vol. 42 ›› Issue (5): 305-308.doi: 10.11896/j.issn.1002-137X.2015.05.062

• 图形图像与模式识别 • 上一篇    下一篇

基于边缘先验模型的运动去模糊

赵志刚,陈莹莹,赵 毅,张维忠,吕慧显,潘振宽   

  1. 青岛大学信息工程学院 青岛266071,青岛大学信息工程学院 青岛266071,青岛大学信息工程学院 青岛266071,青岛大学信息工程学院 青岛266071,青岛大学信息工程学院 青岛266071,青岛大学信息工程学院 青岛266071
  • 出版日期:2018-11-14 发布日期:2018-11-14
  • 基金资助:
    本文受国家自然基金(61170106),山东省科技计划项目(2012YD01058)资助

Motion Deblurring Based on Edge Prior Model

ZHAO Zhi-gang, CHEN Ying-ying, ZHAO Yi, ZHANG Wei-zhong, LV Hui-xian and PAN Zhen-kuan   

  • Online:2018-11-14 Published:2018-11-14

摘要: 从单幅运动模糊图像复原出清晰的图像,一直是数字图像处理领域中富有挑战的问题。基于边缘先验模型和小波分析提出了一种运动模糊退化图像的复原算法。在去模糊之前,对图像进行预处理,将噪声去除,用冲击滤波器增强边缘,并采用canny边缘检测获取清晰边缘作为先验模型,以此估计模糊核;然后在紧小波框架系统下,将清晰图像的稀疏性最大化,采用改进的分裂Bregman方法求解最优化问题,最终得到清晰的图像。实验结果表明,相对于传统的盲复原算法,提出的方法可以有效地去除运动模糊。

关键词: 非盲图像复原,运动模糊,边缘先验模型,分裂Bregman方法,紧小波框架

Abstract: How to restore clear images from degraded images is a challenging problem in the field of digital image processing.This paper proposed a restoration algorithm for motion blurred image based on edge prior model and wavelet analysis.We applied a preprocessing procedure before deblurring to remove noise,and then used shock filter to enhance edges and canny edge detection to get clear edges for estimating kernel.After that,we maximized the sparsity of clear image under tight wavelet frame system.Furthermore,the adapted split Bregman method was proposed to solve the optimization problem.Finally,the clear image can be got.The experimental results show that compared with traditional blind restoration algorithms,the proposed method can effectively remove the motion blur.

Key words: Non-blind image restoration,Motion blur,Edge prior model,Split Bregman method,Tight wavelet frame

[1] Chan T F,Wong C K.Total variation blind deconvolution [J].IEEE Trans.Image Processing,1998(1057-7149):370-375
[2] Chen W G,Nandhakumar N,Martin W N.Image motion estimation from motion smear-a new computational model [J].IEEE Trans.Pattern Analysis Machine Intelligence,1996(0162-8828):412-425
[3] Rav-Acha A,Peleg S.Two motion-blurred images are betterthan one [J].Pattern Recognition Letters,2005,26(3):311-317
[4] Fergus R,Singh B,Hertzmann A,et al.Removing camera shake from a single photograph [J].SIGGRAPH,2006,25(3):783-794
[5] Cai Jian-Feng,Liu Chao-qiang,Ji Hui,et al.Blind motion deblurring from a single image using sparse approximation[C]∥IEEE Conference on Computer Vision and Pattern Recognition,2009(CVPR 2009).Miami,FL,2009:104-111
[6] Cho S,Lee S.Fast Motion Deblurring [J].ACM Transactions on Graphics,2009,28(5):145
[7] Wang C,Yue Y,Dong F,et al.NEAS:Nonedge-Specific Adaptive Scheme for Highly Robust Blind Motion Deblurring of Natural Imagess[J].IEEE Transactions on Image Processing,2013,22(3):884-897
[8] Zhang Hui.Bregman iteration method and its applications tosparsity problems [D].HuNan:National University of Defense Technology,2009
[9] Wang Jing,Lv Ke,He Ning,et al.Total Variant Image Deblurring Based on Split Bregman Method[J].Acta Electronicsa Sinica,2012,40(8):1503-1507
[10] Zhang Xu-ming,Xu Bin-shi,Dong Shi-yun,et al.Adaptive median-weight meam mix filter [J].Optical Technology,2004,30(6):652-655,9
[11] Osher S,Rudin L I.Feature-oriented image enhancement using shock filters [J].SIAM Journal on Numerical Analysis,1990,27(4):919-940
[12] Tomasi C,Manduchi R.Bilateral Filtering for Gray and Color Images[C]∥Proceedings of International Conference on Computer Vision.Bombay,1998:839-846
[13] Qi Shan,Jiaya Jia,Agarwala A.High-quality Motion Deblurring from a Single Image[J].ACM Transactions on Graphics (SIGGRAPH),2008,27(3):1-10

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!