Image Fusion Based on Improved Multi-seed Region Growing and Dual Channel SCM.

A fusion algorithm based on Improved Multi-seed Region Growing combined with Dual Channel Spiking Cortical Model (DCSCM) is proposed to improve the problem of incomplete and unstable target extraction in traditional infrared and visible image fusion algorithm. Firstly, the source infrared image and the source visible light image are decomposed into their own high and low frequency subband coefficients by using the Nonsubsampled Shear-let Transform (NSST). Then, the target information of the source infrared image is extracted by using the improved multi-seed region growing, and the low-frequency fusion coefficient is obtained by comparing the information entropy. The high-frequency fusion coefficient is obtained by using DCSCM. Finally, the fused image is obtained by inverse NSST transform. Simulation results show that the fused image obtained by the algorithm proposed in this paper is rich in infrared target information and detailed texture information, and has good visual effect. In objective evaluation, the fused image has higher standard deviation, information entropy, mutual information and edge preservation coefficient, it has obvious advantages over other methods in subjective and objective aspects.