摘要
在分析了图像经小波分解后边缘信息容易丢失等问题的基础上,给出了小波形态学图像融合改进方法。为了提高图像质量,提出改进主要在低频系数选择算法上,采用形态学灰度形态梯度边缘检测算子对低频子图像进行边缘检测,能够更好的保存原图边缘细节信息;对高频系数选择采用基于绝对值最大的原则。对三组图像进行实验,实验表明采用改进方法使得局部对比度信息得到更好的保留,提高了分辨率,具有一定的实用性。
Based on the analyses of the edge information lost after wavelet decomposition of images,proposed an improved image fusion algorithm based on morphology wavelets transform.The improvement is mainly the selection algorithm of the low-frequency coefficient.In the selection of low-frequency component coefficient,the operator of the morphological gradient edge detection is used to detect the edge,so that the original edge details are preserved.For the high-frequency images,we select those that have maximal absolute values and verify the consistency of these coefficients.This algorithm is tested by three images.The experiment shows that the proposed method retained the local contrast information much better and enhanced the resolution.
出处
《计算机仿真》
CSCD
北大核心
2011年第2期288-291,共4页
Computer Simulation
关键词
图像融合
数学形态学
形态梯度检测算子
小波变换
频率系数
Image fusion
Mathematical morphology
Operator of the morphological gradient detection
Wavelets transform
Frequency coefficient
