Experiment Of Infrared Polarization Imaging

Traditional imaging detection technology focus on intensity and spectral information of the target.,But now the nature background is more complex and the targets are camouflaged and hidden, so under this circumstances, we need to study new imaging method to detect the target under complex background. Infrared polarization imaging based on infrared imaging and polarization imaging is capable to obtain target polarization characteristics which is different from intensity and phase, thereby add the target information dimension. Infrared polarization imaging is not only able to obtain infrared images, but the target polarization information, therefore can enhance and improve target detection and identification capability.Based on the polarization theory, according to Fresnel formula, Kirchhoff theory and Planck’s law, this paper studied the theory of target infrared polarization properties, especially researched on target polarization characteristics of the reflected radiation and spontaneous emission polarization characteristics by simulation. Then introduced polarization representation method: Stokes vector, researched on the principles of polarization imaging techniques and currently the main polarization imaging methods, provided a basis to build an infrared polarization imaging platform. Secondly, according to the model of target polarization characteristics, Systematically analyzed the impact of the observation angle, surface roughness, refractive index, wavelength and other factors when performing infrared polarization imaging, the relevant factors were analyzed by simulation and verified by experiment; Then built a long-wave infrared polarization imaging experimental platform and obtained polarization status of indoor and outdoor scenes by infrared polarization imaging platform and visible imaging system, the results show that the degree of polarization could improve target background contrast scene and significantly suppress the grayscale value of background. Because different direction polarization image of infrared polarization imaging system does not match, proposed a image registration method. By image registration, the clarity and information entropy of the degree of polarization image is increased. Considering the degree of polarization of the target scene increasing the contrast but reducing brightness and resolution of the scene, proposed a polarization-based image fusion method based on non-subsampled shear waves conversion, the method combines the advantages of the light intensity image and the degree of polarization image. The result shows that fused image can improve the information entropy and average gradient of the target scene, at the same time ensure high target background contrast.