摘要
针对红外和可见光图像各自特征,提出了基于小波变换多分辨率的不同规则图像融合方法。首先对原图像进行多级小波分解,产生多频带小波系数,利用像素的局部区域方差、平均梯度及掩膜滤波等方法对不同频带小波系数计算融合权重,最后将小波融合系数逆变换实现图像的融合处理。实验结果表明,融合图像特征显著,易于观察。
According to the respective characteristics of infrared and visible image, a method of multi-rule image fusion based on wavelet transform is proposed. First, the original image is decomposed by multistage wavelet. Second, a multi-band wavelet coefficient is produced. And then, the wavelet coefficients of different frequency band are calculated by local variance of pixels, average gradient, mask filter, and so on. Finally, the fusion image is achieved by the wavelet coefficient reverse transform. The experimental results show that the characteristics of fusion image are distinct and easy to be observed.
出处
《电讯技术》
北大核心
2012年第11期1778-1782,共5页
Telecommunication Engineering
基金
国家自然科学基金资助项目(61162007)
广西壮族自治区教育厅科研项目(201106LX202
201106LX174)~~
关键词
红外图像
可见光图像
多规则图像融合
小波变换
性能评估
infrared image
visible image
multi-rule image fusion
wavelet transform
performance evaluation
