全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

复数小波域的高斯尺度混合模型图像降噪

DOI: 10.11834/jig.20080504

Keywords: 图像降噪,二元树复数小波变换,高斯尺度混合模型,贝叶斯最小均方估计

Full-Text   Cite this paper   Add to My Lib

Abstract:

由于复数小波变换具有近似平移不变性和良好的方向选择性,因此适用于图像去噪。为了取得更好的降噪效果,提出了一种基于复数小波的高斯尺度混合模型降噪算法。该算法首先对自然图像的复数小波系数建立统计模型,即将位于相邻位置和尺度的系数邻域建模为一个高斯尺度混合模型;然后用该模型对子带系数进行贝叶斯最小均方估计,以达到降低噪声的目的。由于这一模型很好地利用了复数小波系数幅值尺度间和尺度内的相关性,因此可以取得较好的降噪效果。实验结果表明,该算法无论从峰值信噪比还是从主观视觉上都优于一些传统的降噪算法。

References

[1]  Donoho D L.Denoising by soft-thresholding[J].IEEE Transactions on Information Theory,1995,41(3):617~627.
[2]  Portilla J,Strela V,Wainwright M J,et al.Image denoising using Ganssian scale mixtures in the wavelet domain[J].IEEE Transactions on Image Processing,2003,12 (11):1338~1351.
[3]  Cransc M S,Nowark R D,Baraniuk R G.Wavelet-based signal processing using hidden Markov modes[J].IEEE Transactions on Signal Processing,1998,46 (4):886~902.
[4]  Portilla J,Strela V,Wainwright M,et al.Adaptive Wiener denaising using a Ganssian scale mixture model in the wavelet domain[A].In:Proceedings of 8th IEEE International Conference on Image Processing[C],Thessaioniki,Greece,2001:37~40.
[5]  Andrews D,Mallows C.Scale mixtures of normal distributions[J].Journal of the Royal Statistical Society.Series B(Methodological),1974,36(1):99~102.
[6]  Kingsbury N.Image processing with complex wavelets[J].Philosophical Transactions on Royal Society London A,1999,35 (160):2543~2560.
[7]  Coifman R R,Donoho D L.Translation-invariant denaising[A].In:A.Antoniadis and G.Oppenheim,Eds:Wavelets and Statistics,Lecture Notes[C],New York,NY,USA:Springer-Verlag,1995:125 ~ 150.
[8]  Chang S G,Yu B,Vettedi M.Spatially adaptive wavelet thresholding with context modeling for image denoising[J].IEEE Transacitans on Image Processing,2000,9(9):1522~1531.
[9]  Simoncelli E P,Adelson E.Noise removal via Bayesian wavelet coring[A].In:Proceedings of IEEE International Conference on Image Processing[C],Lausanne,Switzerland,1996:379 ~ 382.
[10]  Sendur L,Selesnick I W.Bivariate shrinkage functions for wavelet-based denaising exploiting interscalee dependency[J].IEEE Transactions on Signal Processing,2002,50 (11):2744 ~ 2756.
[11]  Wainwright M J,Simoncelli E P.Scale mixtures of Gausaians and the statistics of natural images[J].Advances in Neural Information Processing Systems,2000,12 (1):855 ~ 861.
[12]  Kingsbury N.The dual-tree complex wavelet transform:a new efficient tool for image restoration and enhancement[A].In:Proceedings of European Signal Processing[C],Rhodes,Greece,1998,319~322.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133