摘要
提出了一种基于第二代曲波变换的图像融合方法。首先对源图像进行曲波变换,粗尺度系数采用加权平均的融合规则,细尺度系数采用基于区域标准差的融合规则,然后对细尺度系数进行一致性校验,最后进行图像重构得到融合图像。数值实验结果表明,该方法有效,性能优于基于小波变换的图像融合方法。
A method of image fusion based on the second generation curvelet transform is proposed.Firstly,the source images are decomposed by curvelet transform,then the coarse coefficients are fused with weighted mean fusion rule,and the detail coefficients are fused with fusion rule based on region standard deviation.Secondly,the coherence of the detailed coefficients is verified.Finally,the fused coefficients are reconstructed to obtain fusion image.The numerical experiment results show that the method is effective,and its capability is better than the image fusion method based on wavelet transform.
出处
《现代电子技术》
2011年第15期94-96,共3页
Modern Electronics Technique
关键词
曲波
图像融合
小波
区域标准差
curvelet
image fusion
wavelet
region standard deviation
作者简介
黄昌军 男，1972年出生，陕西黄陵人，讲师，高级程序员，在职研究生。主要研究方向为数字水印、图像处理等。
