Analysis Of Infrared Polarization Imaging Characteristics

Polarization is one of the basic properties of light.Reflection or spontaneous emission on the surface of an object will produce partially polarized light.Polarization characteristics depend on the object's own properties,and there are significant differences in polarization characteristics between different objects.At present,the infrared polarization imaging model is mainly for the polarization model of reflected light,and there is no in-depth study on the infrared spontaneous emission polarization model.Infrared spontaneous emission polarization has broad application prospects in target recognition,geological survey,medical diagnosis and scene recognition,and therefore has important theoretical and practical significance.In this paper,an in-depth study of the polarization characteristics of the infrared target is carried out.The effects of the reflected radiation and spontaneous radiation of the target on the polarization characteristics of the target are comprehensively analyzed and the infrared polarization model is established by combining the two.In the systematic analysis of the influencing factors of infrared polarized imaging,the experimental platform of infrared polarized imaging is designed and established,and an image fusion algorithm is researched and proposed.This article mainly studies the following aspects.(1)Aiming at the diffuse reflection and volume scattering effects produced by the rough surface,the polarization model of the reflected radiation of the rough surface is established.In this paper,based on the Kubelka-Munk(K-M)theory,the diffuse reflection and volume scattering effects on the surface of the object are analyzed,and the diffuse reflection polarization model is established.By adding the diffuse reflection component to the microplane model,an improved polarization bidirectional reflection distribution function model is proposed,and an experimental simulation is established by establishing a reflection polarization model.The simulation results show that the diffuse reflection component will seriously affect the reflection polarization characteristics of the target.(2)Aiming at the infrared band target spontaneous emission will affect the target polarization characteristics,a rough surface spontaneous emission polarization model is established.This paper first analyzes the polarization mechanism of infrared spontaneous radiation on a smooth surface,and performs simulation experiments on smooth aluminum plates and glass.The experiment founAd that the polarization characteristics of spontaneous radiation are greatly affected by the material.Then,based on the concept of the emissivity vector of the black body and the bidirectional reflection distribution function,a mixed radiation emission Stokes vector of rough surface is proposed.Finally,considering the effects of reflected radiation and spontaneous radiation on the polarization characteristics of the target,a complete rough surface infrared polarization model is proposed,and the ratio of reflected radiation to spontaneous radiation is the main factor affecting the degree of infrared polarization.And the infrared polarization measurement environment is constructed to verify the model,and the experimental results are in good agreement with the model results.(3)An infrared polarization imaging experiment platform is designed and built based on the systematic analysis of the influencing factors of infrared polarization.Through the experimental platform,polarized images of the target object in different directions were obtained,and the polarization factors were used to verify and analyze the polarization influencing factors.At the same time,to solve the problems of low gray value and low information entropy of infrared polarization images,an image fusion algorithm based on multi-scale sparse representation and PCNN is studied and implemented.By fusing the polarization image and the intensity image,the algorithm enriches the information entropy of the fused image and improves the average gray value of the image while ensuring the contrast between the target and the background.The experimental results show that The algorithm has a good effect on the fusion of infrared intensity image and polarization image.