一种新的基于压缩感知的多焦点图象融合方法

A Novel Multi-focus Image Fusion Algorithm based on Compressive Sensing

提出一种基于压缩感知和脉冲耦合神经网络的图像融合算法,包括三部分:图像傅里叶变换稀疏表示、测量融合和图像重构。首先,将双通道脉冲耦合神经网络(Dual-PCNN)模型应用到整个算法当中;其次,针对新的融合框架及傅里叶变换的系数特点,提出双星型采样下基于测量值标准差的加权融合方法;最后,通过最小全变分算法重构图像。实验仿真结果证明该方法优于其他基于傅立叶变换的方法。

A novel multi-focus image fusion algorithm based on compressive sensing and the pulse-coupled neural network is proposed,this algorithm consists of three parts： image Fourier transform sparse representation,measurement fusion and image reconstruction. Firstly,the Dual-PCNN（ Dual-Channel Pulse Coupled Neural Network） model is applied in the whole algorithm. Then,aiming at coefficient characteristics ofthe novel fusion frame and the Fourier transform,weight fusion method based on standard deviation of the measured values under the double star sample is proposed. Finally,the fused image is reconstructed by Total Variation algorithm. Experiment simulation results indicate that the proposed algorithm is superior to other existing Fourier transform algorithms.