基于区间双正交小波的图像去噪  

The Noise Removal of Image Based on Interval Biorthogonal Wavelet

作　　者：张彬[1] 王卫卫[1] 尚晓清[1] 宋国乡[1] 

机构地区：[1]西安电子科技大学理学院,西安710071

出　　处：《光子学报》2003年第8期981-984,共4页Acta Photonica Sinica

基　　金：陕西省自然科学基金 (2 0 0 0SL0 2 )资助项目

摘　　要：提出了基于区间双正交小波的图像去噪方法 ,首先用消失矩N和N～ 都是 3的 [0 ,1]区间上的双正交滤波器对含噪图像进行小波变换 ,再用Donoho的通用阈值方法对含有高斯白噪声的加噪图像进行去噪处理 ,仿真结果表明区间双正交小波不仅比Daubechies正交和双正交小波能更好的保留图像的纹理和边缘 ,而且提高了信噪比 .A method for image denoising based on biorthogonal wavelet on an limited interval was proposed. First, DWT was made on the distorted image with a set of biorthogonal wavele filters on the interval and of which the vanishing moments N and N～ are both 3. Then Donoho′s universal thresholding method was applied to reduce the AGWN in the distorted image. Simulation results show that biorthogonal wavelet on an limited interval not only better retain the texture or edges of the image than Daubechies′s biorthogonal and orthogonal wavelet, but also improve the SNR.

关 键 词：区间双正交小波 图像去噪 软阈值 

分 类 号：TN911.73[电子电信—通信与信息系统]