Research On Image Fusion Technology Based On Polarization

The detection of targets in complex background environment,especially the identification of stealth,camouflage,faint and false targets,has become an important research topic in today's battlefield for intelligence acquisition and real-time reconnaissance.From the practical applications of imaging,the general imaging technology can only detect the intensity of light information,and because the light in the process of transmission,will be affected by the atmospheric haze scattering and reflection effect,so the image acquired by imaging system become blurred,that it is difficult to distinguish the information of the target in the background.As a new type of detection technology,polarization imaging detection can not only detect the intensity information of the target,but also obtain parameters such as spectrum,polarization state and spatial geometry other than the intensity.Analyzing the polarization information can invert the intrinsic property of the target,and then through the image fusion technology,you can extract hidden or more difficult to find the target information,the goal highlighted from the background,in order to achieve the ability to obtain intelligence on the battlefield.In this paper,the polarization state of polarized light,the mathematical description of polarized light and the characterization method of polarized optical system are discussed in detail.According to the overall design scheme of time-division full polarization imaging detection system,polarization based on Stokes vector and Mueller matrix Parameter Measurement of Imaging Detection System.Using the polarization imaging system,different experimental scenes and experimental conditions were designed,and a great deal of polarization image acquisition was completed.The experimental results show that the polarized parameter image can reflect the rich information of the object well,while the degree of polarization and angle of polarization images can highlight the texture and edge information of the target.In order to make better use of these polarization information,a polarization image fusion algorithm based on two-dimensional discrete wavelet transform,a polarization image fusion algorithm based on discrete multiwavelet transform,and a polarization image fusion algorithm based on discrete curvelet transform are proposed.By using Stokes parametric images and the different features of the images with all polarization parameters are taken into account,and the algorithm is further optimized and improved.The image fusion algorithm not only reduces the background information,but also improves the contrast and clarity of the target,which makes it easier to identify the hidden or difficult to distinguish by naked eye.At the end of this paper,the fusion image is evaluated and compared by the edge intensity,average gradient,information entropy and contrast.