High-Quality Computational Ghost Imaging with a Conditional GAN

Abstract

In this study, we demonstrated a framework for improving the image quality of computational ghost imaging (CGI) that used a conditional generative adversarial network (cGAN). With a set of low-quality images from a CGI system and their corresponding ground-truth counterparts, a cGAN was trained that could generate high-quality images from new low-quality images. The results showed that compared with the traditional method based on compressed sensing, this method greatly improved the image quality when the sampling ratio was low.