Research On Image Fusion Technology Based On Visible Light And Infrared Radiation Dual-band Target Detection

Due to the development of sensor technology,the big data analysis information technology mastered by humans has also increased rapidly,including the technology of visible light and infrared image fusion.The visible light and infrared image fusion technology integrates the effective information of the two source images,removes useless redundant information,and solves the problem that the single information is limited by the imaging method,imaging spectrum and natural factors,reducing the target scene.Uncertainty improves the practicability of the system.Therefore,the research on heterogeneous image fusion technology has great practical value.The visible light and infrared image fusion algorithm studied in this paper adopts the pixel-level fusion level and selects the fusion method of multi-scale transformation,in which the multi-scale tool selects non-subsampled shear waves.Compared with other fusion methods,the fusion method based on non-subsampled shearlet can better extract image features.Therefore,based on the non-subsampled shearlet transform,this paper mainly studies two image fusion methods according to the current research status,one is the fusion method of grayscale visible light and infrared images,and the other is the fusion method of color visible light and infrared images.Image fusion method,and integrate the two methods on an image fusion platform interface.The details are as follows:1.Aiming at the problems of low salience of infrared targets in fusion images and insufficient detail information in visible light images,this paper proposes an image fusion algorithm based on non-subsampled shearlet and improved sigmoid function.In the low-frequency part,the fusion weight of infrared low-frequency sub-images is calculated based on the improved sigmoid weight configuration function,thereby determining the fusion ratio of visible light and infrared images;Images are fused.The experimental results show that the algorithm studied in this paper has a significant effect on improving the saliency of the infrared target,making the target and the background have a clear distinction,and retains more visible light detail texture information in the image,so that the fusion image can be The visual effect has been greatly improved.2.Aiming at the problems of low salient target pixel value and unnatural color performance in color fusion images,this paper proposes a fusion algorithm for color visible light and infrared images.The mapping theory combines the method of image gradient energy to formulate the low-frequency fusion strategy of the brightness component of the color visible photon image and the infrared sub-image.The experimental results show that the algorithm studied in this paper can not only improve the pixel value of the infrared salient target,but also restore the background color in the color visible light image,and improve the overall visual effect of the fusion image.3.For the above two algorithms,this paper designs a visible light and infrared image fusion platform,uses the Qt cross-platform interface application framework and the Open CV machine vision library to develop software in the Visual studio 2017 environment,and designs and implements a dual-light image fusion system.The overall framework of the two algorithms is better reflected.The experimental results show that the fusion platform perfectly encapsulates the two algorithms and has a good interactive interface to realize the fusion of visible light and infrared images.