Polarized computational ghost imaging in scattering system with half-cyclic sinusoidal patterns

Abstract

Computational Ghost Imaging (CGI) can reconstruct the scene images through the correlation algorithm between the preset illumination patterns and the intensities of the detection bucket detector. In the actual transmission process, the scattering system in the imaging path can affect the imaging quality of the CGI. The half-cyclic sinusoidal-pattern based CGI (HCSP-CGI) can recover the scene images at low sampling rates (SRs). Based on the Monte Carlo (MC) model, we propose and construct a reflective polarization system based the HCSP-CGI in fog environment and put forward a method that combines polarization information with the HCSP-CGI called the PHCSP-CGI. When the difference in reflectivity between target and background is small, the difference in polarization characteristics between target and background can help CGI to remove the interference of stray light. The imaging performance of the PHCSP-CGI at different SRs and scattering distances (SDs) are analyzed. Experimental results show that high-contrast scene can still be obtained, providing new applications for object recognition in scattering system.