Improvement Of Polarization Image Quality Based On Multiple Polarization Parameters

It is a research topic of significant practical meaning to improve the target identification ability of an optical imaging system in a complicated environment,and such topic is always in the spotlight.The polarimetric imaging techniques can utilize the multiple-dimensional polarization information to well identify the targets under complicated conditions.In the present paper,several key issues in the polarimetric imaging techniques are investigated,with the purpose to optimize the target identification ability of a polarimetric imaging system in a complicated environment.The main work of this paper is as follows.The testing on metal surface damage is performed.It can be seen from the experiments that the images of the degree of polarization(Do P)and of the angle of polarization(Ao P),under a condition of non-uniformed illumination,represent stronger target identification ability than the traditional intensity images.In the paper,with regard to the defects in the prior method for characterizing Ao P images,a method for characterizing Ao P images based on a 3D HSI color space is proposed,lowering the characterization noise and also making the image better conform to the appearance features of an object.Though the polarimetric imaging techniques can decrease the influence of nonuniform illumination,there is still a relatively strong noise in the image under the condition of low illumination,thus seriously decreasing the contrast of the polarized image.For this problem,in the present paper,based on the traditional method for optimizing the contrast of the polarimetric imaging and according to the inherent correlation of the multiple-dimensional polarized images,a method for noise reduction of polarized images is proposed.It is proved in theory that the Fisher value of an image can be improved by 100% when it is optimized by a regular tetrahedron measuring matrix.It can be seen from the experiments that the present method can effectively lower the noise of the polarized image and improve the image contrast.Finally,in the present paper,polarimetric dehazing method in the environment of scattering media is studied.For the strong scattering environment,the Do P of the scattered light,as a key parameter in the typical dehazing method,is investigated.Based on the polarization maintaining characteristic of the circularly polarized light,a method for estimating the Do P of the scattered light is proposed,not only improving the effect of the polarimetric dehazing method,but also simplifying the dehazing algorithm.In particular,though the polarimetric dehazing techniques can achieve an acceptable effect of dehazing,it can obtain only the 1D intensity image of the scene,rather than the polarization information of the scene.For this defect,in the present paper,a polarimetric dehazing method is proposed in which the polarization information of the target in the scattering media is obtained.In the present method,the Stokes vector is used to describe the complete polarization states of the target light and the scattered light in the scattering environment.Based on such physical model,the multipledimensional polarization parameters are used to obtain the 3D information of the polarized images of the scene.Then an image fusion method conforming to the physical meaning of polarization is designed,and the color image is obtained to characterize the polarization information of the scene,thus significantly improving the target identification ability of the polarimetric dehazing techniques.