摘要
针对低对比度图像融合时会造成图像细节模糊等缺陷,提出一种基于加权双正交自适应小波变换的图像融合算法。取小波系数局部模极大并进行自适应加权修正以融合高频系数。采用对2幅原图像的低频系数加权自适应的方法对低频系数进行融合。对多聚焦图像和多光谱彩色图像分别采用计算图像信息熵和均方根误差、计算图像平均梯度的方法对融合的性能进行评价。实验结果表明,采用该融合规则得到的融合图像具有良好的融合效果。
An image fusion algorithm with weighted biorthogonal self adaptive wavelet transform is proposed, which can make up for defects that there are faintness of image details in image fusion of lower contrast image. The adaptive method of wavelet coefficient local model maximum which is weighted is used to fuse the high frequency component, and the syncretic adaptive method is also chosen in the course of fusing low frequency coefficient. The capability of multi-focus image fusion is evaluated by calculating image information entropy and Root Mean Square Error（RMSE）. The capability of multi-spectrum image fusion is evaluated by calculating mean grads of image. Experimental results show that the fusion rule of the method is more effective.
出处
《计算机工程》
CAS
CSCD
北大核心
2009年第11期228-230,共3页
Computer Engineering
基金
广西教育厅科研基金资助项目(200707LX196)
广西工学院自然科学基金资助项目(院科自0704102)
关键词
小波变换
图像融合
图像信息熵
均方根误差
wavelet transform
image fusion
image information entropy
Root Mean Square Error（RMSE）
