Aiming at the physical characteristics of infrared and visible imaging sensors, a novel image fusion algorithm based on the second generation Curvelet transform was proposed. Firstly, the fast discrete Curvelet transform was performed on the original images respectively to obtain the subband coefficients at different scales and in various directions. Then for low frequency subband coefficients, the fusion weights were determined by the target characteristics of infrared image and the detail information of visible image; while for high frequency subband coefficients, a fusion rule based on local-region energy matching was employed. Finally, the fusion results were obtained through the inverse Curvelet transform. Experimental results show that the proposed algorithm can effectively integrate important information from infrared and visible images, and the obtained results are better than those of pyramid-based or wavelet-based algorithms.