Research On Joint Probability Drop Shadow For Selection Of Bimodal Infrared Image Fusion Algorithm

Infrared intensity image is imaging based on the difference of infrared radiation energy of the target,while infrared polarization image is imaging based on the polarization property of the target.As a result,they exhibit significant differences in terms of brightness,edge detection,and texture details.Due to their unique imaging advantages,they can complement each other and improve the reliability and application value of the images.However,existing applications of bimodal infrared image fusion often fail to achieve the desired results due to their reliance on users’ prior knowledge.Therefore,this thesis investigates the different types of differential features in bimodal infrared images and uses them as a basis for analysing the contribution of differential features in image fusion.A fusion effectiveness distribution analysis is carried out for multiple attributes of the disparity features to better evaluate the fusion effectiveness of these features in different situations.Subsequently,the joint possibility drop shadow is constructed by means of possibility distribution synthesis,enabling adaptive selection of the best fusion strategy to complete the bimodal IR image fusion,thereby improving the quality of the fusion.The main work of this thesis is as follows:(1)Construction of dual-mode infrared image differential feature set,fusion algorithm set and evaluation index set: This thesis begins with the imaging principle of dual-mode infrared images and examines the benefits of dual-mode infrared intensity imaging and polarimetric imaging.It also investigates the roles and properties of various characteristics.Building upon this,a difference feature set is developed to quantify the complementary information between the two types of images in terms of grayscale,edge,and texture.Through experimental analysis,the validity and effectiveness of the difference feature set are demonstrated.(2)Construction of fusion efficiency of difference features of dual-mode infrared images:This thesis summarizes a large number of image fusion algorithms and image fusion quality evaluation methods,analyzes the fusion requirements and functions of dual-mode infrared images,and determines the fusion algorithm set and fusion quality evaluation index set for the experiments in this paper.It also analyzes the role of fusion effectiveness,compares common fusion effectiveness measurement methods,constructs a reasonable fusion effectiveness,and verifies its validity and effectiveness.(3)Possibility distribution and joint possibility drop shadow construction of the fusion effectiveness of difference feature: The possibility distribution of fusion effectiveness and joint possibility drop shadow function were constructed from the perspective of two-dimensional space by analyzing the theory of possibility distribution,and the optimal fusion algorithm was selected by combining the frequency of difference feature distribution.From the perspective of three-dimensional space,the position information of image features is retained,and the fusion efficiency possibility distribution and joint possibility drop shadow surface are constructed to realize the selection of the optimal fusion algorithm.