Research On Fusion Algorithm Of Infrared And Visible Light Image Under Low Light

As the most direct and direct information type,image information has attracted wide attention from scholars.In particular,the image information fusion of infrared and visible light,because of the imaging characteristics of infrared and visible light,their image information has a good complementarity.It can effectively excavate and integrate the feature information of the image,highlight the infrared target,enhance the understanding of the scene and so on.It is of great value in the fields of medicine,military,monitoring,search and rescue and so on.With the development of information fusion technology,a variety of fusion methods have been proposed.These methods can be divided into three categories: substitution method,neural network method and multi-scale transformation method.For example,independent component analysis(ICA)and principal component analysis(PCA)are used to extract main information and apply to image fusion.However,these algorithms are easy to lose some details or slow processing speed,which affects the image fusion effect when extracting the main information of the source image.The fusion method of visible and infrared images is studied in this paper.The fused images can reflect the target objects more comprehensively under the condition of low light.Image fusion can be divided into three parts: image decomposition,coefficient fusion and image reconstruction.In this paper,the algorithm is improved from two aspects of image decomposition and reconstruction and fusion rules.The main research contents include the following aspects: the fusion of infrared image and visible light image combined with two.First,the imaging characteristics of infrared and visible light sensors are discussed.The weight model and image fusion model are established for the system analysis of infrared and visible light under the light of light,and the evaluation index and evaluation basis of the image fusion are introduced and discussed in detail.Secondly,the preprocessing method of infrared image is studied and analyzed.From the three aspects of image registration,image smoothing and enhancement of target contrast,better pretreatment methods are compared through experimental analysis.The fusion effect of infrared and visible images is enhanced,and the observability of fused images is enhanced.Then,the existing discrete wavelet transform(DWT),principal component analysis(PCA)and non down sampling Contourlet transform(NSCT)fusion algorithm are summarized and summarized.According to the transformation theory of these algorithms,the simulation test is done.Compared with the simulation test results,the shortcomings are pointed out,and the theoretical method and the contrast data are provided for the improved algorithm.Finally,two improved fusion algorithms are proposed,the method of the algorithm is analyzed,the model system is established,the experimental steps are summarized,and the experimental results are compared with other algorithms.The experimental results prove that the two infrared and visible image fusion algorithms are all effective,and can more effectively retain the target information and spatial structure information in the source image,and integrate the image more effectively.The algorithm is effective for clarity and information content.To sum up,the fusion technology of infrared and visible light images is studied.In view of the characteristics of infrared and visible light images,the fusion effect of the existing algorithms is studied,and two improved image fusion algorithms are proposed.Simulation results show that the proposed algorithm has better fusion effect.