浏览全部资源
扫码关注微信
哈尔滨工程大学 自动化学院,黑龙江 哈尔滨,150001
收稿日期:2012-11-22,
修回日期:2013-02-02,
网络出版日期:2013-06-20,
纸质出版日期:2013-06-15
移动端阅览
朱齐丹 孙磊 蔡成涛. 应用自适应权值矩阵的图像复原[J]. 光学精密工程, 2013,21(6): 1592-1597
ZHU Qi-dan SUN Lei CAI Cheng-tao. Image restoration using adaptive weight matrix[J]. Editorial Office of Optics and Precision Engineering, 2013,21(6): 1592-1597
朱齐丹 孙磊 蔡成涛. 应用自适应权值矩阵的图像复原[J]. 光学精密工程, 2013,21(6): 1592-1597 DOI: 10.3788/OPE.20132106.1592.
ZHU Qi-dan SUN Lei CAI Cheng-tao. Image restoration using adaptive weight matrix[J]. Editorial Office of Optics and Precision Engineering, 2013,21(6): 1592-1597 DOI: 10.3788/OPE.20132106.1592.
由于自然图像先验模型的概率密度函数的非高斯性(稀疏性的)会导致图像复原的最优化函数不再是凸的,传统意义上的最大后验估计已不能很好地求得最优估计解,因此本文提出利用自适应权值矩阵来解决这一问题。进行图像复原时,首先利用自然图像先验模型有效地抑制振铃效应,然后利用基于自适应权值矩阵的共轭梯度算法来解决由于稀疏先验模型导致的最优化函数非凸的问题。权值矩阵可根据上一次的迭代结果进行更新,并能够纠正图像在上一次迭代过程中局部区域导数估计的错误。实验结果显示,利用本文方法得到复原图像的峰值信噪比(PSNR)为36.131 6,优于其它算法。最后,用本文方法对全景图像进行复原,得到了很好的复原效果,证明了本文方法的实用性和有效性。
Because natural image filter response probability model is often a non-Gaussian form(sparse)
it leads to the optimization problem of image restoration to be a non-convex and the optimal estimation solution can not be obtained by traditional maximum a posterior estimation. Therefore
this paper proposes a image restoration algorithm based on adaptive weight matrix to solve the problem. With image restoration
a prior model for the natural image is used to restrain the ring effect effectively and the conjugate gradient algorithm based on adaptive weight matrix is used to solve the problem of the non-convex optimization function due to the sparse prior. The weight matrix updates according to the last iteration result and is able to correct the error of the local image derivative estimation in the last iteration process. Experiments show that Peak Signal to Noise Ratio( PSNR)gotten by proposed algorithm is 36.131 6
better than that from other algorithms. Finally
the panoramic image is restored by proposed method and the good results are also obtained
which demonstrates that the algorithm proposed is practical and effective.
FERGUS R,SINGH B,HERTZMANN A,et al.. Removing camera shake from a single photograph[J]. ACM Trans.Graph, 2006, 25(3):787-794.[2]KRISHNAN D,FERGUS R.Fast image deconvolution using hyper-Laplacian priors[C]. NIPS,2009:1-9.[3]HACOHEN Y,FATTAL R,LISCHINSKI D. Image upsampling via texture hallucination[C]. Proceedings of the IEEE International Conference on Computational Photography (ICCP), 2010:1-8.[4]CHO T S,ZITNICH C L,JOSHI N,et al.. Image Restoration by matching gradient distributions.[J].IEEE Tran. on Pattern Analysis and Machine Intelligence.2012,34(4):683-694.[5]温博,张启衡,张建林. 应用自解卷积和增量Wiener滤波实现迭代盲图像复原[J]. 光学 精密工程,2011,19(12):3049-3055.WEN B, ZHANG Q H,ZHANG J L. Realization of iterative blind image restoration by self deconvolution and increment Wiener filter [J].Opt. Precision Eng., 2012,34(4):683-694. (in Chinese)[6]ROTH S,BLACK M J. Fields of experts[J]. International Journal of Computer Vision,2009, 82(2):205-229.[7]WEISS,FREEMAN W T. What makes a good model of natural images [C]. CVPR, 2007:1-8.[8]CHO T S,JOSHI N,ZITNICK C L,et al.. A content-aware image prior [C].CVPR, 2010:169-176.[9]SCHMIDT U,GAO Q,ROTH S. A generative perspective on MRFs in low-level vision [C]. CVPR,2010:1751-1758.[10]邹谋炎.反卷积与信号复原[M]. 北京: 国防工业出版社, 2001:35-39.ZOU M Y. Deconvolution and Signal Recovery [M].Beijing: Nation Defense Industry Press,2001:35-39.(in Chinese)[11]TAPPEN M F,LIU C,ADELSON E H,et al.. Learning Gaussian conditional random fields for low-level vision [C].CVPR, 2007:1-8.[12]MEER P. Emerging Topics in Computer Vision[M]. Prentice Hall,2004.[13]ROTH S,BLACK M J. Fields of experts:A frame of experts for learning image priors[C].CVPR,2005:860-867.[14]SCHMIDT U,SCHELTEN K,ROTH S. Bayesian Deblurring with Integrated Noise Estimation [C].CVPR,2011:2625-2632.[15]曾吉勇, 苏显渝. 双曲面折反射全景成像系统[J].光学学报,2003, 23(9):1138-1142.ZENG J Y, SU X Y. Hype rboloidal catadioptric omnidirectioal imaging system[J].Acta Optica Sinica, 2003,23(9):1138-1142. (in Chinese)
0
浏览量
581
下载量
1
CSCD
关联资源
相关文章
相关作者
相关机构