基于子波变换的图像多通道阈值降噪法  

Multi-Channel Threshold Image Denoising Based on the Wavelet Transfom

作　　者：殷瑞祥[1] 张晓明[1] 

机构地区：[1]华南理工大学电子与信息工程学院,广东广州510640

出　　处：《华南理工大学学报（自然科学版）》2001年第5期89-93,共5页Journal of South China University of Technology(Natural Science Edition)

摘　　要：由于子波变换良好的信号能量集中性 ,近年来子波变换域的图像降噪受到人们重视 .子波变换域阈值去噪法因处理过程简单而得到广泛应用 .图像信息经过子波变换后在子波域分布的情况与子波基的选取有关 ,因此 ,不同的子波基在信噪分离上具有互补作用 .本文利用这种互补作用提出了两种子波变换域多通道阈值降噪方法 .根据噪声对图像影响与图像自身特征有关 ,以及子波变换域各子带子波系数之间的相关性 ,还提出了一种自适应的通道阈值设定方法 .实验结果表明 ,提出的算法在保持处理过程简单的同时 ,达到了更好的滤除噪声效果 .Since the wavelet transform possesses good property of energy centralization, the image denoising based on wavelet is paid great attention to recently. With simple processing progress, the thres_hold denoising is widely used. As allocation of the image information in wavelet transform field is dependent on the wavelet base used, different wavelet bases are mutual in signal_noise departing. This paper proposes two multi_channel image denoising methods in wavelet transform field. Since the noise_interference with the image depends on the characteristic of the image and the wavelet coefficients in different subband are correlative, this paper also presents an adapted adjusting method of channel thres_hold. The experimental results indicate that the proposed algorithm achieves better denoising performance and also possesses a simple processing progress.

关 键 词：子波变换 图像去噪 多通道阈值降噪法 子波基 阈值滤波法 信噪比 

分 类 号：TN911.73[电子电信—通信与信息系统]