关键词: 图像融合, 多尺度几何分析, 非下采样轮廓波变换, 脉冲耦合神经网络

Abstract: To overcome the defects of image fusion algorithm using Wavelet transform and simple fusion rules，an improved image fusion algorithm based on pulse coupled neural networks and non-subsampled Contourlet transform is put forward in this paper.The input registered images are decomposed by NSCT.The Contourlet domain coefficients in different scales and directions are fused by improved fusion rules based on PCNN.The fusion image is acquired by recomposing the fused coefficients.Experiments show that this scheme has acquired promising results both in the subjective vision aspect and the objective assessment parameters.

Key words: image fusion, Multiscale Geometric Analysis（MGA）, nonsubsampled Contourlet transform, Pulse Coupled Neural Networks（PCNN）