基于噪声方差确定非线性扩散除噪声的最优停止时间

刘鹏; 刘定生; 李国庆; 李景山

doi:10.3724/SP.J.1146.2008.01279

基于噪声方差确定非线性扩散除噪声的最优停止时间

doi: 10.3724/SP.J.1146.2008.01279

基金项目:

国家863计划项目资助课题

计量
- 文章访问数: 3802
- HTML全文浏览量: 86
- PDF下载量: 771
- 被引次数: 0
出版历程
- 收稿日期: 2008-10-06
- 修回日期: 2009-04-21
- 刊出日期: 2009-09-19

The Selection of Optimal Stopping Time Based on Synchronous Iteration of Noise and Image in Diffusion Image De-noise

摘要

摘要: 该文采用非线性扩散进行图像除噪声并在这个计算框架下提出利用噪声方差选择最优停止时间的方法。在利用非线性扩散进行图像除噪声时，每次迭代平滑掉的图像的方差大于平滑掉的噪声的方差时，迭代应该停止。为了在除噪声过程中正确地估计噪声的方差，该文构造一幅纯噪声图像跟实际的观测图像同步进行迭代计算，并把纯噪声图像的方差作为图像中噪声方差的估计值来辅助计算最优停止时间。针对非线性扩散的各项异性，提出了能够保持两种噪声同步变化的特殊的规整化项。新的规整化项在迭代纯粹噪声图像时使用，这样确保每次迭代都可以保持合成噪声与实际图像噪声的统计特性相一致。实验证明新的算法可以非常有效地选择合适的停止时间。
- 图像处理; 最优停止时间; 扩散滤波; 噪声方差
Abstract: This paper de-noises in image by diffusion filter. And a method of finding optimal stopping time is proposed. The criterion of stopping time is that the variance of image smoothed out is bigger than the variance of noise smoothed out. In order to estimate the variance of noise in iteration correctly, a pure synthesis noise as an image is synchronously iterated with the observation image in iteration, and the variance of pure noise image is taken as the estimation of the variance of noise in estimated image. According to the anisotropy of the regularization, a novel regularization term that can ensure the synchronous changing of the synthesis noise and real noise was proposed in this article. The new regularization term is put into use only in iteration of pure noise image, and the similarity of statistical properties between real noise and synthesis noise can be kept in iteration. Experiment confirms the effectiveness of proposed method to select optimal stopping time.

HTML全文

参考文献(1)

Perona P and Malik J. Scale-space and edge detection usinganisotropic diffusion[J].IEEE Transactions on PatternAnalysis and Machine Intelligence.1990, 12(7):629-639[2]Rudin L, Osher S, and Fatemi E. Nonlinear Total Variationbased noise removal algorithms[J].Physical D.1992, 60(1-4):259-268[3]Teboul S, Blanc-Feraud L, Aubert G, and Barlaud M.Variational approach for edge-preserving regularization usingcoupled PDE's[J].IEEE Transactions on Image Processing.1998, 7(3):387-397[4]芦碧波, 郭志昌. 非局部偏微分方程去噪模型[J]. 吉林大学学报(理学版), 2007, 45(6): 935-936.Lu Bi-bo and Guo Zhi-chang. Nonlocal partial differentialequation model for image denoising[J]. Journal of JilinUnicersity(Science Edition), 2007, 45(6): 935-936.[5]张红英, 吴亚东, 吴斌. 基于变分PDE 的非线性数字混合滤波器[J]. 计算机辅助设计与图形学学报, 2007, 19(9):1089-1093.Zhang Hong-ying, Wu Ya-dong, and Wu Bin. Nonlineardigital hybrid filter based on variational PDE[J]. Journal ofcomputer-Aided Design Computer Graphics, 2007, 19(9):1089-1093.[6]邵文泽, 韦志辉. 局部几何结构驱动的图像插值放大及超分辨率复原[J]. 中国图象图形学报, 2008, 13(7): 1235-1243.Shao Wen-ze and Wei Zhi-hui.Local geometry driven imagemagnification and super-resolution[J]. Journal of Image andGraphics, 2008, 13(7): 1235-1243.[7]邵文泽, 韦志辉. 基于广义Huber-MRF 图像建模的超分辨率复原算法[J]. 软件学报, 2007, 18(10): 2434-2444.Shao Wen-ze and Wei Zhi-hui. Super-resolution imagerestoration based on Huber-MRF modle[J]. Journal ofSoftware, 2007, 18(10): 2434-2444.[8]Weickert J. Coherence-enhancing diffusion of color images[J].Image and Vision Computing.1999, 17(3):201-212[9]Mrzek P and Navara M. Selection of optimal stoppingtime for nonlinear diffusion filtering[J].International Journalof Computer Vision.2003, 52(2/3):189-203[10]Gilboa G, Sochen N, and Zeevi Y Y. Estimation of optimalPDE-based denoising in the SNR sense[J].IEEETransactions on Image Processing.2006, 15(8):2269-2280[11]Charbonnier P, Blanc-Feraud L, Aubert G, and Barlaud M.Deterministic edgepreserving regularization in computedimaging[J].IEEE Transactions on Image Processing.1997,6(2):298-311[12]Tschumperl D. Vector-valued image regularization withPDEs: A common framework for different applications[J].IEEE Transactions on Pattern Analysis and. MachineIntelligence, 2005, 12(7): 629-639.

施引文献

资源附件(0)

访问统计