基于双边滤波和双树复小波的图像去噪算法  

Image Denoising Algorithm Based on Bilateral Filtering and Dual-Tree Complex Wavelet

作　　者：曹剑中[1] 周祚峰[1] 唐垚[1] 郭敏[1] 王浩[1] 

机构地区：[1]中国科学院西安光学精密机械研究所,西安710119

出　　处：《光子学报》2010年第9期1712-1715,共4页Acta Photonica Sinica

基　　金：中国科学院"西部之光"人才培养计划"西部博士资助项目"资助

摘　　要：提出了基于空域双边滤波和双树复小波变换的图像去噪算法.该算法使用双树复小波变换对含噪图像进行多尺度和多方向的分解,对各个高频方向子带使用带有方向窗的局部维纳滤波算法进行去噪.在重构过程中,对每一个尺度上重构得到的低通图像使用空域的双边滤波算法进一步的去除噪声.实验结果表明本文提出的图像去噪算法获得了明显的去噪性能改善.An image denoising algorithm based on bilateral filtering and dual-tree complex wavelet transform is proposed.The noisy image is decomposed into multi-scale and multi-directional subbands by dual-tree complex wavelet transform,and each high-pass directional subband is denoised by local Wiener filtering with directional windows.During the reconstruction,each low-pass constructed image is further denoised by the bilateral filtering in spatial domain.The experiment results show that the proposed image denoising algorithm achieves significant improvement of denoising performance.

关 键 词：双边滤波 双树复小波 局部维纳滤波 图像去噪 

分 类 号：TN911.73[电子电信—通信与信息系统]