Research On Infrared And Visible Image Fusion Based On Non-Subsampled Shearlet Transform

Infrared and visible images have good complementary characteristics,infrared and visible image fusion technology has important application value in many fields.The infrared image targets are prominent,and it is easy be disturbed by the external environment during imaging,but infrared image lack object details,and its image contrast and spatial resolution are low;on the contrary,the visible image has high contrast,spatial resolution and rich details,but it is easy to be affected by the external environment during imaging,and it is difficult to detect hidden targets in the scene.Therefore,it is of great significance to research the infrared and visible image fusion technology.In this paper,infrared and visible image fusion algorithms are researched based on non-subsampled shearlet transform(NSST),three novel image fusion algorithms based on multi-scale transform are proposed.The main work of this paper is as follows:1.Aiming at the traditional image fusion algorithms based on multi-scale transform have problems in the edge information missing and the target feature being not prominent enough et al,a novel infrared and visible image fusion algorithm based on optimized pulse coupled neural network(PCNN)and region feature guided rule is proposed.The NSST is used to decompose the source image into low-frequency and high-frequency parts,and the traditional PCNN model is optimized to fuse the low-frequency part.At the same time,use the feature of image,such as region energy,improved spatial frequency and region variance matching degree et al,an adaptive threshold of region variance matching degree and new regulator factors are proposed,thus the region feature guided rule is constructed to fuse the high-frequency part.Experimental results show that the fusion performance of the proposed algorithm is better,the infrared target is prominent and edge contour is clear on the fused image.2.Aiming at some image fusion algorithms based on saliency detection have problems in targets being not prominent and low contrast et al,a novel infrared and visible image fusion method by combining the improved saliency detection and NSST is proposed.The improved maximum symmetric surround algorithm is used to extract the saliency map of infrared image,then the improved gamma correction method is used to enhance it,and the visible image is enhanced by homomorphic filtering.Next,infrared image and the enhanced visible image are decomposed into low-frequency and high-frequency parts by NSST,and the saliency map is used to guide the lowfrequency part fusion.At the same time,the rule of maximum region energy selection is used to guide the high-frequency part fusion.Experimental results show that the proposed method has prominent infrared targets,more background details,high overall contrast and definition on the fused image.3.In order to solve the shortcomings of current traditional image fusion algorithms,such as the targets are not prominent enough and many texture details are lost,a novel infrared and visible image fusion algorithm based on Gaussian fuzzy logic and adaptive dual-channel spiking cortical model(ADCSCM)is proposed.The source image is decomposed into low-frequency and high-frequency parts by NSST.Then,combine with new sum of Laplacian and Gaussian fuzzy logic,and set dual thresholds to guide the low-frequency part fusion.At the same time,the fusion rule based on ADCSCM is used to guide the high-frequency part fusion.Experimental results show that the proposed algorithm has better subjective visual effect,which can effectively highlight infrared targets and retain more texture details.