Infrared image and visual image fusion algorithm based on NSCT and improved regional cross entropy

In order to overcome the defect of multi-scale decomposition and reconstruction tools for wavelet transform and that of Contourlet Transform tool which does not have the translation invariance, multi-scale, multi-directional properties and the translation invariance of NSCT are chosen to transfer the source images in frequency domain. An improved regional cross entropy fusion rule is used for the lowpass subband component which reflects approximate contents. The fusion method for the regional average gradient is used for the bandpass directional subbands component which reflects detail properties of image. Finally, the fusion image is reconstructed through an inverse transform of NSCT. Experimental results are given by comparing with those of the methods based on wavelet transform or Contourlet transform. These results show that by using this algorithm, the mutual information between the infrared and visual images can be fused more organically, the details of fusion image are richer, the target is more prominent, and the edge information are reappeared better. Strong support for the follow-up information analysis and extractive ability of the images are provided.