基于双树复小波变换的图像融合

Image Fusion Based on Dual-Tree Complex Wavelet Transform

为弥补小波变换的不足,采用基于双树复小波变换的方法进行图像融合。在对源图像进行双树复小波分解后,综合考虑了图像的高低频成分,对低频成分采用区域平均能量择大的融合规则,高频成分采用局部标准差结合选择及加权平均的策略,最后采用双树复小波逆变换重构图像。经多聚焦及多光谱图像的融合实验及主客观考核,结果表明:融合后的图像更清晰地反映了源图像的细节信息,融合质量相对较高。

To make up for the lack of wavelet transform,this paper uses a method based on dual- tree complex wavelet transform（ DT- CWT） for image fusion. After decomposing the source image by DT- CWT,this algorithm considers both low- frequency and high- frequency components of the image. The low- frequency components are fused by the choose- max strategy with the local average energy; the high- frequency components use the local- standard- deviation to being the basis of selective and weighted- averaging fusion method. Finally the inverse DT- CWT is used to reconstruct the image. Experiments of multi- focus images and multi-spectral images and subjective and objective assessment results reveal that the fused image clearly reflects the details of the image,and that the fusion quality is relatively high.