基于正交小波变换的图像去噪算法研究  

Image Denoising Algorithm Based on Orthogonal Wavelet Transform

作　　者：覃焕昌[1] 滕政胜[1] 

机构地区：[1]广西百色学院物理与电信工程系,广西百色533000

出　　处：《通信技术》2009年第1期290-291,311,共3页Communications Technology

摘　　要：提出了一种基于正交小波变换的图像去噪方法,首先利用离散小波对图像信号按Mallat算法进行分解,然后采用软阈值与小波重构的算法进行去噪。深入研究了小波变换中的图像分解与重构的Mallat算法,详细介绍了正交小波变换中阈值的选取,并进行了实验研究。实验结果表明,该方法可以有效去除噪声,并保留了图像细节部分的有用信息。This paper tells of new image denoising method based on orthogonal wavelet transform. Based on the Mallat algorithm, the image signal is first decomposed through the wavelet. Then, the denoising algorithm by soft threshold and the wavelet reconstruction is used to remove the noise. Mallat algorithms of image decomposition and reconstruction are studied deeply, the method for selecting threshold in wavelet transform is described in detail. Experimental results show that the proposed scheme can remove the noise effectively and preserve more useful details in image denoising.

关 键 词：正交小波 图像降噪 MALLAT算法 软阈值 

分 类 号：TP371[自动化与计算机技术—计算机系统结构]