摘要
提出了基于多小波变换的图像处理方法,该方法以多小波变换为基础,在一次多小波分解与重构之间完成双谱段图像处理。首先进行多小波变换,将变换系数进行软阈值收缩消去噪声;然后根据图像中需增强的信息,选择增强系数进行子带增强;最后提出一种新的自适应权值融合规则,采用这个规则融合变换系数,进行小波重构得到处理后的单幅图像。实验表明,这种方法不仅能提高图像的视觉效果,增强源图像的边缘信息,而且能很好地将源图像中对电晕检测有用的信息融合在一起,提高电晕检测系统的定位精度。
With combination of the multiwavelet threshold shrinkage, subband enhancement and image fusion, a image processing method accomplished by means of discrete multiwavelet transform was presented. In this method, Multi Wavelet Transform(MWT) is the first step and the MWT coeffi cients are denoised by soft threshold multiwavelet shrinkage. Then subband enhancement is used to enhance the edge related coefficients. A new adaptive weight average image fusion rule was proposed to merge the coefficients and acquire fused coefficients. The experimental results show that the pro- posed image processing method can produce visually acceptable image and reduce noise while the source image is enhanced. This method also can fuse details of input images and improve the locating precision of the corona detection system.
出处
《光学精密工程》
EI
CAS
CSCD
北大核心
2006年第4期714-719,共6页
Optics and Precision Engineering
基金
教育部博士点基金资助项目(No.20010183)
关键词
多小波变换
电晕检测
图像融合
多小波收缩
子带增强
Multi Wavelet Transform(MWT)
corona detection
image fusion
multiwavelet shrinkage
subband enhancement