Research On Super-resolution Reconstruction Algorithm And Realization Technology Of Infrared Micro-polarization Detection

With the development of science and technology,the performance of infrared imaging systems is getting higher and higher,the ability to recognize targets is getting stronger and stronger,and their applications are becoming more and more extensive.But at the same time,the traditional infrared imaging technology is also facing the challenge of infrared countermeasure technology.Infrared countermeasure technology uses artificial infrared interference and infrared camouflage and other methods,so that the traditional infrared imaging system cannot accurately and effectively detect the target.Infrared polarization imaging technology can improve the detection ability of camouflage and stealth targets.Infrared polarization imaging technology can not only obtain the infrared intensity information of the target,but also obtain the polarization information of the target.Since different objects generally show different polarization states,Therefore,the obtained polarization information can show some outlines information of the target,so that the infrared polarization imaging technology can recognize the camouflage target more effectively than the traditional infrared detection technology.In this paper,the Stokes vector is used to describe the polarization state of the target.According to the polarization data in 4 directions,the polarization information of the target can be calculated using Stokes vector.The infrared polarization detection device used in this article is an uncooled infrared polarization focal plane array.This array is processed with a grating of an angle in front of each pixel.The grating of the entire array has a total of 4 polarization directions and is regularly distributed,so the polarization data in 4 directions can be acquired at the same time.However,this grating angle distribution also leads to a reduction in the resolution of the four polarization angle data.If the polarization information is directly calculated,a low-resolution polarization information image will be obtained.Therefore,it is necessary to use a super-resolution reconstruction algorithm to process the polarization data and finally obtain a high-resolution image.For the problem that the resolution of the polarization information is reduced,this paper first obtains the infrared polarization data of the target by building an infrared polarization detection imaging system,and then analyzes the grating angle distribution law of the uncooled infrared polarization focal plane array,and proposes three reconstruction methods of polarization information image,these three methods are super pixel calculation and super-resolution reconstruction method,misalignment calculation method,super-resolution reconstruction to restore the original data method.Through the evaluation and comparison of the image quality reconstructed by the three methods,it is found that the polarization information image obtained by the third reconstruction method has more details and the reconstruction effect is better.This paper then proposes an improved super-resolution reconstruction algorithm for infrared micro-polarized images.This algorithm is based on the third reconstruction method,super-resolution reconstruction to restore the original data method,which uses a combination of bicubic interpolation and gradient-based interpolation to achieve the enhancement of edge information.After evaluating the quality of the reconstructed image and comparing with the previous reconstruction methods,it can be concluded that this improved algorithm is superior to the previous three reconstruction methods,which can effectively enhance the edge information and improve the image quality.And this improved algorithm is easy to implement and can realize the real-time processing of the algorithm on hardware.