Infrared dual-band image fusion with simplified pulse coupled neural network and visual saliency map in nonsubsampled shearlet transform domain

As an effective way to integrate complementary information in multisensor detection system, image fusion technology has been widely used in robotic vision, medical diagnosis and safety monitoring. At the same time, the dual band infrared detection system has been widely used in the field of guidance and detection.Because dual-band/multi-band infrared detection has the characteristics of wide detection range and multi-target radiation information. Therefore, there is an urgent need of a fusion of the dual-bands infrared images. In order to obtain better image quality, infrared dual-frequency image fusion technology is used to synthesize different radiation information of target and background.In this paper, a new infrared dual-band image fusion with simplified pulse coupled neural network(PCNN) and visual saliency map(VSM) Framework in nonsubampled shearlet domain (NSST) is proposed. In the proposed method, first, the sours images are decomposed into base parts and multiscale and multidirection representations in NSST domain. Then,base parts are fused by VSM fusion approach. For the high-frequency bands are fused by a Simplified pulse coupled neural network model. Finally, the final image is reconstructed by inverse NSST. As a result, the fused image details will be presented more naturally, which is more suitable for human visual perception. The experimental results demonstrate that evaluation quality of the fused images is improved by comparing three objective evaluation factors with three popular fusion methods.This technology is of great significance to the development of image field.