基于方向对比度和局部方差的双树复小波图像融合算法

Image fusion via dual-tree complex wavelet transform based on directional contrast and local variation

本文提出了一种在双树复小波(Dual-Tree ComplexW avelet Transform,DT-CWT)变换域下的融合策略,将小波域的方向对比度推广到复小波变换域,并将其与局部方差结合形成新的融合规则,实验结果表明该融合规则可以更好的突出融合图像的对比度信息,更有效的保留纹理和细节等局部特性。

A novel multi-resolution image fusion method via dual-tree complex wavelet transform （DT- CWT） based on directional contrast and local variation is presented. Firstly, the source images are decomposed by dual-tree complex wavelet transform. Then the coefficients for the fused image can be ob- tained by the novel fusion rules based on maximum directional contrast and local variation. Lastly, the fused image is reconstructed by inverse dual-tree complex wavelet transform. The experimental results show that this method can enhance details and eontrast information of fused image.