A fusion algorithm based on contourlet transform and improved pulse coupled neural network was proposed. By using the contourlet transform, the input images were firstly decomposed into a number of sub-images with various scale and directional features. Then, based on the IPCNN, a fusion rule in the contourlet domain was given. The fused coefficients could be generated by the IPCNN based fusion rule and the fused image was obtained by performing the inverse contourlet transform to the fused coefficients. The proposed algorithm was successfully applied for the visible image and infrared image fusion. The simulation results confirm the validity of the proposed method.