Technique for infrared and visible image fusion based on non-subsampled shearlet transform and spiking cortical model

Fusion of infrared and visible images is an active research area in image processing, and a variety of relevant algorithms have been developed. However, the existing techniques commonly cannot gain good fusion performance and acceptable computational complexity simultaneously. This paper proposes a novel image fusion approach that integrates the non-subsampled shearlet transform (NSST) with spiking cortical model (SCM) to overcome the above drawbacks. On the one hand, using NSST to conduct the decompositions and reconstruction not only consists with human vision characteristics, but also effectively decreases the computational complexity compared with the current popular multi-resolution analysis tools such as non-subsampled contourlet transform (NSCT). On the other hand, SCM, which has been considered to be an optimal neuron network model recently, is responsible for the fusion of sub-images from different scales and directions. Experimental results indicate that the proposed method is promising, and it does significantly improve the fusion quality in both aspects of subjective visual performance and objective comparisons compared with other current popular ones.