Polarization Characteristic Extraction Based on Multi-Angle

With the more and more extensive application of polarization in target detection and other fields, the traditional images of degree of polarization and angle of polarization can no longer meet the needs of highlighting target characteristics. Therefore, we propose a calculation method of polarization characteristic image, which reflects the difference of polarization characteristics of targets. The process is as follows: firstly, the parallel and vertical polarization component images with polarization detection angles of 0°-90° and 90°-180° are calculated. Secondly, the global parallel and vertical component images are determined by comparing the gradient sizes of the parallel and vertical component images in two angle intervals, and the multi-angle polarization characteristics (MPC) are obtained by orthogonal difference between them. We use contrast (C), average gradient (AG), image entropy (E) and peak signal-to-noise ratio (PSNR) to compare S 1 , S 2 , DOLP, AOP and MPC respectively. The experimental results show that MPC has more advantages in subjective visual perception and objective indicators. Experiments show that our MPC image can be used for target detection.