Nonsubsampled contourlet transform（ NSCT） not only has translation invariance

but also has sufficient redundancy information

and it can more effectively extract direction information from source image

so that the fused image is more consistent with human visual characteristics. A new image fusion algorithm for visible-infrared image based on improved region energy and image self-similarity in NSCT domain is proposed. The two source images are decomposed into low-frequency subbands and high-frequency subbands by NSCT. As for the low frequency subband coefficients

an regional energy adaptive weighting image fusion algorithm is presented; while as for the high-frequency subband coefficients

based on self-similarity. Fusion coefficients are reconstructed by inverse NSCT to reconstruct to form fusion image. Compared with some existing algorithms

the proposed fusion algorithm results has obvious advantages in the human eye subjective visual standards and objective evaluation criteria.