Two-layer Pulse Coupled Neural Network Model for Image Fusion

Pulse coupled neural network is an effective model for image fusion. Single-channel PCNN is used to calculate fusion coefficients, while multi-channel PCNN is used to fuse image directly. This paper proposed a novel two-layer Pulse Coupled Neural Network, called Cascade Pulse Coupled Neural Network (CPCNN). The proposed model considers coefficients calculation as well as image fusion. The first layer of CPCNN is the part of coefficients calculation, which contains m single-channel PCNNs. The second layer of CPCNN is the part of image fusion, which contains a multi-channel PCNN. Compared with the existing PCNN fusion model, CPCNN is more effective and flexible. Experimental results showed the better performance of CPCNN in both visual effect and objective evaluation criteria.