关键词: 遥感图像融合, Canny算子, 非下采样Contourlet变换, 粒子群优化算法, 结构相似度

Abstract: In order to significantly improve the spatial resolution of the fused image and maximally fuse the spectral information of multi-spectral images, it presents a Particle Swarm Optimized（PSO） remote sensing image fusion method by combining Canny operator with Nonsubsampled Contourlet Transform（NSCT） in this paper. Firstly, the edge information of the panchromatic image is distinguished to the non-edge by Canny operator on the basis of IHS transform, and the edge of the panchromatic image is enhanced by fusing the panchromatic image and [I] component of multi-spectral image according to the characteristics of edge distribution. Then, the enhanced panchromatic image and [I] component of multi-spectral image are respectively decomposed by NSCT, the method of selective weighted summation is used in the lowpass sub-band, the PSO is performed to search the best threshold [p] and the fusion rules of regional structure similarity are used in the highpass sub-band. Finally, the fused image is reconstructed by inverse transform of NSCT and IHS. The simulation experiment results show that the algorithm can maintain a good balance between the detail information and spectral information of the original images.

Key words: remote sensing image fusion, Canny operator, nonsubsampled Contourlet transform, particle swarm optimization, structural similarity