提出了一种新的基于非亚采样Contourlet变换的PET/CT图像融合方法。通过非亚采样金字塔(NSP)和非亚采样方向滤波器组(NSDFB)实现对图像的多尺度多方向分解。在融合处理中,对分解图像所得的高频及低频系数,根据不同分解面的特性,采用不同的加权规则进行融合。该方法既保留了Contourlet变换方法的多分辨率特性,又具有平移不变性。实验结果表明,相对于传统的拉普拉斯金字塔变换法、小波变换法和Contourlet变换法等,取得了更佳的融合效果

A novel PET/CT images fusion method based on nonsubsampled contourlet transform(NSCT) is presented, which uses the nonsubsampled pyramids(NSP) and the nonsubsampled directional filter banks(NSDFBs) to decompose the source images in a multi-direction way. Like the contourlet transform method, the presented method can obtain multi-scale and multi-direction images with good multi-resolutions, and it is also shift-invariant. The high and low frequency-coefficients are obtained by decomposing the source images with nonsubsampled contourlet transformation. Then, according to the different coefficients characters, these coefficients are syncretized by using different fusion algorithms of weighted average. The experimental results show that the new method makes a better fusion result compared with the traditional transform methods.