In order to solve the problem that the rules of traditional image fusi on cannot make full use of global image features,this article used pulse coupled neural network (PCNN) model into the Curvelet transform image fusi on,made the Support Value (SPV) of the characterization sub band image characteristics as the external excitation input to stimulate PCN N model,at the same time,after considering the correlation of information between low and high frequency sub band of the Curvelet transform,s et PCNN model parameters (strength and range of the connection) to change with the characteristics of the low frequency sub band ima ges adaptively,and using the first ignition timing of each neuron in the PCNN model to structure the significant measure of fusion rule.Us ing the PCNN model to simulate the biological characteristics of human visual neural system,and use the global coupling features to analyze a nd process the fusion of the source image intelligently,so as to improve the overall effect of the fusion images.The experimental results show t hat,when PCNN is used to participate in selecting detail coefficients,due to the global coupling characteristics and pulse synchronizati on feature,it can use the global information of sub band image better.