Restoration of Single pixel imaging in atmospheric turbulence by Fourier filter and CGAN

As a novel computational imaging (CI) method, single pixel imaging (SPI) can obtain the spatial information of the target object with only a single-pixel detector, especially the SPI with Fourier basis has higher imaging efficiency and quality. However, Fourier single-pixel imaging (FSI) will still be affected in atmospheric turbulence, imaging will be unstable, and image quality will decrease. This paper first uses the phase screen method to simulate atmospheric turbulence, and then in the FSI process, the Fourier coefficients are processed to filter noise in the Fourier domain to reduce the impact of turbulence on imaging. Then we use the conditional generative adversarial network (CGAN) of deep learning methods to further improve the clarity and quality of the restored image.