Research On Infrared And Visible Image Fusion Algorithm Based On Multiscale Decomposition F I I

Infrared and visible light image fusion is a major research hotspot in the field of image fusion.It has important applications in military,aviation,and security surveillance.The purpose of the fusion of infrared and visible light images is to effectively synthesize the significant target information in the infrared image and the rich texture information in the visible light image.The existing fusion algorithms have the problems of introducing more artifacts,low target saliency,loss of texture details,and long fusion time.In response to the above problems,the following researches have been carried out in this thesis:1.Aiming at the problem of not prominent salient targets and many artifacts in image fusion,an image combining fast and adaptive bidimensional empirical mode decomposition(FABMED)and improved saliency detection Fusion Algorithm.The source image is decomposed into a base layer and a detail layer by FABEMD.The improved saliency detection algorithm and guided filtering are used to generate the fusion rules of each layer.Experimental results show that the fusion algorithm better retains the saliency of the target in the source image,improves the contrast of the fused image,and introduces fewer artifacts during the fusion process.2.To solve the problem of losing detailed information in the source image in the fused image,an image fusion algorithm based on an improved adaptive dual-channel pulse-coupled neural network(ADC-PCNN)and saliency detection.The source image is decomposed into a base layer and a detail layer by FABEMD.Using ADC-PCNN and saliency detection to generate fusion rules for each layer.Experimental results show that the fused image of this algorithm can retain rich details and improve the quality of fusion.3.Aiming at the problem that image fusion takes a long time and has limitations in practical applications,a fast image fusion algorithm combining two-scale decomposition and improved saliency detection is proposed.The mean image is used to decompose the source image into base layer and detail layer.The improved saliency detection and guided filtering are used to generate the fusion rules of each layer.Experimental results show that the algorithm has low time consumption and good fusion quality,which reflects the feasibility.