基于灰色关联度的提升小波域图像融合算法  

Image Fusion Methodin Lifting Wavelet Domain Based on Gray Correlation Degree

作　　者：王丽[1] 苗凤娟[1] 陶佰睿[2] 张景林[3] 刘文礼[1] 

机构地区：[1]齐齐哈尔大学通信与电子工程学院,黑龙江齐齐哈尔161005 [2]齐齐哈尔大学计算中心,黑龙江齐齐哈尔161005 [3]齐齐哈尔大学网络中心,黑龙江齐齐哈尔161005

出　　处：《电视技术》2015年第1期15-18,共4页Video Engineering

基　　金：黑龙江省自然科学基金项目(F201332);黑龙江省教育厅科学技术研究项目(12531774)

摘　　要：针对多聚焦图像,提出了一种基于提升静态小波变换(Lifting Stationary Wavelet Transform,LSWT)和灰色关联度相结合的图像融合算法。首先将待融合的两幅图像分别进行提升静态小波变换,低频分量采用平均法融合。其次针对低频分量和各个方向的高频分量,分别计算以每个像素为中心的方块的能量和均值。然后按照一定的融合规则,分别得到高频分量融合后的值。最后进行提升小波逆变换,得到融合后的图像。实验结果显示,该方法融合效果较好,优于传统算法。A image fusion method based on lifting stationary wavelet transform and gray correlation is presented for multi-focus image. Firstly,the two fused images are decomposed by LSWT. The mean method is used to fuse the low-frequency components. Secondly, for the low-frequency components and high-frequency components in three directions ,energy and mean are calculated for each pixel as the center of the square. Then, according to some fusion rules,three high-frequency fused values are obtained respectively. Finally, by performing the inverse LSWT,the fusion image is obtained. Experimental results show that this fusion method is better than traditional method.

关 键 词：提升静态小波变换 灰色关联度 图像融合 能量 均值 

分 类 号：TN911.73[电子电信—通信与信息系统] TP391.41[电子电信—信息与通信工程]