基于多小波变换的图像去噪技术  

Image denoising with multiwavelets

作　　者：贾杏托[1] 王成儒[1] 

机构地区：[1]燕山大学,河北秦皇岛066004

出　　处：《计算机工程与应用》2010年第19期204-206,237,共4页Computer Engineering and Applications

摘　　要：多小波变换同时拥有正则性、正交性、紧支性、对称性等单小波不能同时满足的优点,近年来在图像处理领域受到了广泛关注。利用多小波变换后系数之间存在的大量的相关性的特点,提出了一种新的小波阈值方法。通过仿真实验,证明了该方法能有效地去除噪声。Multiwavelet transform can meet the merits at the same time,such as regular and orthogonal,compact support,but scalar wavelet can not.Therefore,it attracts a lot of concern in the field of image processing.Based on the dependencies between neighboring coefficients,a new weighted multiwavelet matrix thresholding rule is proposed.And the experiment proves that it is an effective way to remove noise.

关 键 词：多小波 图像降噪 预处理 加权 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]