Polarized Passive Millimeter Wave Image Post-processing Technology

Different characteristics of millimeter wave radiation polarization of different targets is of great significance for target detection and target recognition in passive millimeter wave imaging technology.In this paper,polarimetric passive millimeter wave(PMMW)images are taken as the research object,and relevant post-processing techniques are studied.Firstly,based on the difference of polarization characteristics of target millimeter-wave radiation,a polarimetric PMMW image fusion algorithm based on two-scale decomposition is proposed.The fusion rules are designed according to the visual saliency and the image fusion experiment is carried out simultaneously.The image fusion performance was evaluated both subjectively and objectively.Then,a polarimetric PMMW image fusion algorithm based on Nonsubsampled Contourlet Transform(NSCT)is improved,and a saliency map detection method based on the characteristics of polarized images is designed for improving the fusion rule.Compared with other multi-scale transform fusion algorithms,the proposed algorithm is more effective.Finally,a polarization information display method for PMMW images is studied to compensate for the shortage of polarization information in a single grayscale image.A group of polarimetric PMMW images are converted to HSV color space by threshold segmentation and linear scaling,and then the Pseudo-colorized images are generated.By improving the threshold segmentation,a polarization information display method based on bat algorithm is proposed,which realizes the adaptive multi-threshold segmentation of images,and the effectiveness of the improved algorithm is verified.Furthermore,in order to enable the polarimetric PMMW image to possess the color information of the real scene,research on the Pseudo-colorization algorithm of polarization information based on color migration is carried out.And for the lack of detail,an improved color migration algorith is proposed.