Fast restoration algorithm model for membrane diffraction imaging

Abstract Membrane diffraction imaging is a crucial technology for lightweight and large-diameter space-based telescopes. However, the image quality of membrane diffraction imaging system is deteriorating sharply owing to the multiple diffraction orders of diffraction elements and wavefront distortion of the imaging beam. This study successfully determined the optimal solutions of the imaging model parameters of Fresnel zone plates (FZPs) using the stochastic parallel gradient descent (SPGD) algorithm. Subsequently, we proposed a fast restoration algorithm based on the deconvolutions of the FZP imaging model. We conducted image sampling and restoration experiments on the FZP imaging system. The results demonstrated a significant improvement in the contrast and gradient of the restored image, leading to enhanced texture and detail in the images.