基于多小波收缩与子带增强的图像去噪方法  

Image Denoising Based on Multiwavelet Shrinkage and Subband Enhancement

作　　者：王欣[1] 庞云阶[1] 

机构地区：[1]吉林大学计算机科学与技术学院,长春130012

出　　处：《仪器仪表学报》2004年第z3期380-383,共4页Chinese Journal of Scientific Instrument

基　　金：教育部博士点基金资助项目(20010183)。

摘　　要：边缘特征是图像最为有用的高频信息,因此在图像去噪的同时,应尽量保留图像的边缘特征,基于这一思想,提出了多小波阈值收缩与子带增强相结合的图像去噪方法。该方法以多小波变换为基础,将变换后的多小波系数分为噪声相关系数和边缘相关系数,对变换系数进行软阈值多小波收缩消去噪声相关系数;阈值收缩是非线性变换,对图像边缘有平滑作用,因此该方法提出在阈值收缩后进行线性的子带增强,增强边缘相关系数。实验表明与单一的阈值收缩方法相比,该方法不但保留了图像的边缘特征,而且提高去噪图像的峰值信噪比,实验结果优于普通的阈值收缩方法。Edge information is the most important high frequency information of an image.Therefore it should try to maintain more edge information in the process of denoising.A new image denoising method based on multiwavelet threshold shrinkage and subband enhancement is proposed.Multiwavelet transform is the first step,the MWT coefficients can be divided into two categories:the coefficients associated with noise and the coefficients associated with edges.The coefficients associated with noise are reduced by soft threshold multiwavelet shrinkage.But this procedure is a non-linearity transformation.It causes the edge smoothness.So the subband enhancement function to enhance the edge related coefficients is introduced.The experimental results show that compared with the wavelet threshold shrinkage methods, our denoising method can retain as much as possible the important signal features and increase PSNR.This method performs better than commonly used threshold shrinkage method.

关 键 词：去 噪 多小波变换 多小波收缩 子带增强 

分 类 号：TH7-55[机械工程—仪器科学与技术]