基于曲波系数相关性的去噪算法

被引:6
作者
王文波
羿旭明
费浦生
机构
[1] 武汉大学数学与统计学院
关键词
曲波变换; 贝叶斯估计; 概率密度函数(PDF); 图像降噪;
D O I
10.16136/j.joel.2006.12.026
中图分类号
TN911.73 [图像信号处理];
学科分类号
摘要
在利用曲波对图像进行分解的基础上,提出了一种考虑不同尺度分量的曲波系数间是相互关联的系数概率密度函数(PDF)模型,不同尺度的系数选择依赖其父层系数的大小,从而保证在去除噪声的同时,保留图像的纹理边缘等细节信息,然后对图像进行重构。实验表明,该方法能有效地去除噪声,同时较好地保留了图像的纹理和边缘等细节,而且对图像质量有进一步的改善。
引用
收藏
页码:1519 / 1523
页数:5
相关论文
共 12 条
[1]  
The curvelet trans-formfor i mage denoising. Strack J L,Candes E J,Donoho D L. IEEE Trans on I mage Pro-cessing . 2002
[2]  
I mage fusion method based on sationary multi wavelet transform. ZHANG Xue-shuai,,PAN Quan,ZHAO Yong-qiang,et al. Journal of Optoelectronics·Laser . 2005
[3]  
Edge-preservingi mage denois-ing method using curvelet transform. XIAO Xiao-kui,LI Shao-fa. Journal on Com-munications . 2004
[4]  
I mage reconstruction by the wavelet transform applied to aperture synthesis. Strack J L,,Bijaoui A,Lopez B,et al. Astronomy and Astrophysics . 1994
[5]  
Singularity detection ad process-ing with wavelets. Mallat S G,Hwang WL. IEEE Transactions on Information Theory . 1992
[6]  
Theories andApplications of Wavelets. CHENG Li-zhi,WANG Hong-xia,LUO Yong. . 2004
[7]  
Orthonormal ridgelets andlinear singularities. Donoho D L. SIAM Journal on Mathematical Analysis . 2000
[8]  
Multiscale wavelet based segmentation of extended target in complex environment. KONG Gang,,ZHANG Qi-heng. Journal of Optoelectronics·Laser . 2004
[9]  
The si mulation of generalized Gaussian distribution random sequence. ZHANG Qiang,MA Run-nian,XU Jin. Journal of Xidian University(Natural Science) . 2001
[10]  
Denoising method of remote sensing i mage based on the wavelet transform. HOU Bo. . 2002