摘要
提出一种各分解层上的不同频率分量采用不同的融合算子的双正交小波函数图像融合的新方法。首先将参加融合的两幅图像进行双正交小波多尺度分解,再根据纹理信息函数和边缘亮度函数的相关系数,对分解后的高频部分和低频部分采用不同融合算子进行加权融合。仿真结果表明,该算法与传统的图像融合算法相比较,能得到清晰度更高,偏差指数更小的融合图像。
A new image fusion method based on biorthogonal wavelet that different frequency component on each decomposition level uses different operator is proposed. First, images involved in fusion is decomposed by biorthogonal multi-resolution wavelet.Then different fusion operators are used to high frequency images and low frequency images according to correlation coefficient between texture information function and edge luminance function.The simulation result indicated that the algorithm is able to obtain fused images of higher clarity and lower deviation index compared with traditional methods.
出处
《四川理工学院学报（自然科学版）》
CAS
2008年第4期74-76,共3页
Journal of Sichuan University of Science & Engineering(Natural Science Edition)
基金
四川省教育厅青年基金项目(07ZB150)
关键词
双正交小波
图像融合
融合算子
biorthogonal wavelet
image fusion
fusion operator
