波原子纹理图像阈值算法

刘国军; 冯象初; 张选德

doi:10.3724/SP.J.1146.2008.00595

波原子纹理图像阈值算法

doi: 10.3724/SP.J.1146.2008.00595

刘国军^{①②;冯象初},
冯象初,
张选德

基金项目:

国家自然科学基金(60872138)资助项目

计量
- 文章访问数: 3606
- HTML全文浏览量: 110
- PDF下载量: 819
- 被引次数: 0
出版历程
- 收稿日期: 2008-05-15
- 修回日期: 2009-03-30
- 刊出日期: 2009-08-19

Threshold Algorithm of Texture Images with Wave Atoms

摘要

摘要: 该文提出了一个新的依赖于Besov光滑参数和尺度的波原子软阈值纹理图像去噪模型。该模型充分考虑了新的多尺度几何分析工具波原子的优良特性，比如波原子正交基和框架的灵活选取，对振荡纹理图像的稀疏表示，波长和支撑尺寸满足抛物尺度关系等。数值实验表明，新模型比硬阈值和软阈值具有更好的去噪性能，而且随着Besov光滑参数的增加信噪比有明显提高。
- 图像处理;波原子;纹理;Besov空间;光滑参数;阈值;尺度
Abstract: A novel denoising model for texture images is proposed, which is the soft threshold algorithm depending on both the smoothing parameter in Besov spaces and the scales of wave atoms. This model well considers the good properties of new multiscale geometric analysis toolwave atoms, such as the flexible choice of the orthonormal basis and tight frames, sparse representation of the oscillatory texture images, as well as parabolic scaling between wavelength and the size of the essential support. Numerical experiments show that the proposed model not only has a better denoising performance comparing to the hard and soft threshold, but also significantly improves the SNR with the increase of the smoothing parameter in Besov spaces.

HTML全文

参考文献(1)

Donoho D L and Flesia A Gl. Can recent innovations inharmonic analysis explain key findings in natural imagestatistics [J]. Network: Computation in Neural Systems, 2001,12(3): 371-393.[2]焦李成, 谭山. 图像的多尺度几何分析: 回顾和展望[J]. 电子学报, 2003, 31(12A): 1975-1981.Jiao Li-cheng and Tan Shan. Development and prospect ofimage multiscale geometric analysis[J]. Acta ElectronicaSinica, 2003, 31(12A): 1975-1981.[3]Candes E J. Ridgelets: theory and applications[D]. [PH.D.dissertation], Stanford University, 1998.[4]Candes E J. Monoscale ridgelets for the representation ofimages with edges[R]. [Technical Report], Stanford University,1999.[5]Donoho D L. Orthonormal ridgelet and linear singularities[J].SIAM Journal on Mathematical Analysis.2000, 31(5):1062-1099[6]Candes E J and Donoho D L. Curvelets: a surprisinglyeffective nonadaptive representation for objects with edges[C].Curves and Surfaces Fitting, Saint-Malo 1999. Nashville, TN,2000: 105-120.[7]Candes E J and Donoho D L. New tight frames of crveletsand optimal representations of objects with2 C singularities[J].Communications on Pure and AppliedMathematics.2004, 57(2):219-266[8]Pennec E L and Mallat S. Sparse geometric imagerepresentation with bandelets[J].IEEE Transactions onImage Processing.2005, 14(4):423-438[9]Guo K and Labate D. Optimally sparse multidimensionalrepresentation using Shearlets[J].SIAM Journal onMathematical Analysis.2007, 39(1):298-318[10]汪凯斌, 俞卞章, 赵健等. 基于Gabor 小波的无边缘活动围道纹理分割方法[J].电子与信息学报.2007, 29(12):2819-2821浏览[11]Demanet L and Ying L X. Wave atoms and aparsity ofoscillatory patterns[J].Applied and Computational HarmonicAnalysis.2007, 23(3):368-387[12]Ma Jian-wei. Characterization of textural surfaces using waveatoms[J]. Applied Physics Letter, 2007, 90(5): 1-3.[13]Plonka G and Ma Jian-wei. Nonlinear regularizedreaction-diffusion filters for denoising of images withtextures[J].IEEE Transactions on Image Processing.2008,17(8):1283-1294[14]Villemoes L. Wavelet packets with uniform time-frequencylocation[J].Comptes-Rendus Mathematique.2002, 335(10):793-796[15]Rudin L I, Osher S J, and Fatemi E. Nonlinear Totalvariation based noise removal algorithms [J].Physica D.1992,60(1-4):259-268[16]Chambolle A and Lucier B J. Interpreting translationinvariantwavelet shrinkage as a new image smoothing scalespace[J].IEEE Transactions on Image Processing.2001, 10(7):993-1000[17]DeVore R A. Nonlinear approximation[J].Acta Numerica.1998, 7(1):51-150

施引文献

资源附件(0)

访问统计