Group-based sparse representation for Fourier ptychography microscopy

Fourier ptychography microscopy (FPM), which employs alternative projection between spatial and Fourier domains to stitch low-resolution images captured under angularly varying illumination, reconstructs one image with high-resolution and wide field of view (FOV) to bypass the limitation of the space-bandwidth product (SBP) of the traditional optical system. However, system noises such as pupil aberrations, LEDs misalignment and so on, are inevitably introduced in the process of capturing low-resolution images. To address this problem, we propose a new method to insert the Group-based sparse representation (GSR) into the convergence-related metric of FPM as the regularization term in this paper. We have carried out the experiments over both synthetic and real captured images, and the results demonstrate that the proposed method is able to have promising performance while inhibiting the noises efficiently.