In order to solve the problem that the infrared light sensor and the visible light sensor can’t get clear images simultaneously, a nonsubsampled Contourlet coefficient compressive sensing based on infrared and visible image fusion method was proposed. Firstly, the nonsubsampled Contourlet transformation method was used to do multiscale multi directional image decomposition for the infrared and visible image. Each nonsubsampled Contourlet coefficient of them was got. Then, the low-frequency coefficients of the infrared and visible image were fused by the weighted average fusion method. The band pass sub band coefficients by the pseudo random Fourier matrix observations weights fusion method were used. After that, fused band pass sub band coefficients were reconstructed. Finally, image by the inverse nonsubsampled Contourlet transformation were reconstructed. The experiment results showed that this fusion algorithm can get image with clear object and background, and it has the low computational complexity and good fusion effect. 