基于小波变换的图像融合算法研究

为提高不同聚焦图像的融合质量,提出一种基于小波变换的图像融合改进算法. 利用小波变换将多聚焦图像进行多尺度分解,对分解后的高频系数与低频系数采用不同的选择方案,采用3×3滑动高斯窗口分别计算出待融合图像高频部分的相似度矩阵以及边缘能量矩阵,最后采用将结构相似度与图像边缘算子结合的融合算子对图像进行融合. 结果表明,该改进算法极大地抑制了图像融合中振铃效应,有效避免了融合过程中的信息缺失, 融合图像具有更好的视觉效果. 

Study on Image Fusion Algorithm Based on Wavelet Transform

A kind of image fusion algorithm based on wavelet transform was proposed and it is appropriate for fusion of multi-focus images. First, multi-scale decomposition of each source image was gotten through wavelet transform, the selection scheme of high-frequency and low-frequency coefficients was discussed. While choosing high-frequency coefficients, the SSIM matrix and the edge energy matrix were calculated by using slip gauss window, finally, fusion images were achieved according to a fusion operator which combines structural similarity (SSIM) with edge operator. The results show that the improving algorithm greatly inhibits the ringing effect, effectively avoids the information loss in the fusion process and the effect of fused image is better.