In this paper, a new dual-band infrared image fusion for enhancing visibility is proposed. Before proposing the new algorithm, we studied the human visual systems and found that human visual systems have directional characteristics, and the sensitivity of the human eye to the phase angle is higher than the change of the mode. Considering these characteristics of human visual systems, we put forward a dual-band infrared image fusion algorithm based on local window activity measure. First, the mid-wave and long-wave infrared source images are decomposed into the multi-scale and multidirectional subbands by using a nonsubsampled contourlet transform(NSCT) method. Then, the highpass subbands are fused by selecting the maximum absolute operator and the lowpass subbands are fused by using a method based on region energy and region variance. Finally, the image is reconstructed by inverse nonsubsampled contourlet transform and a fused image is obtained. We use a dual-band infrared camera with common optical path to acquire images for the fusion experiment. The method is compared with the gradient pyramid transform and the wavelet transform in fusion effectiveness. From the comparison of the evaluation parameters, we can find that the fusion results of the proposed algorithm are better than other algorithms, and this visibility-enhanced dual-band infrared image fusion algorithm based NSCT is more suitable for human eye observation and understanding.