Adaptive denoising method for Fourier ptychographic microscopy

Abstract

Fourier ptychographic microscopy (FPM) is a recently developed wide-field and high-resolution (HR) imaging technique, reconstructing HR spectrum from a series of low-resolution (LR) images at different illumination angles. Although many significant progresses have been made in FPM in the past few years, imaging noise is still an inevitable problem, which could seriously distort the results recovered using the conventional Fourier ptychography approach without image preprocessing. Generally, before FPM reconstruction, a thresholding denoising method is usually employed to eliminate the noise. However, conventional thresholding denoising algorithms cannot differentiate useful signals from imaging noise effectively, thus these algorithms usually eliminate signals and noise simultaneously. Here we propose an adaptive denoising method for FPM, which takes advantage of the information redundancy in FPM to separate signal from noise during the recovery process without any pre-knowledge about the noise statistics. Simulation and experimental results are presented to evaluate the performance of the proposed method. It is demonstrated that this method can both improve the accuracy and robustness of FPM and relax the imaging performance requirement for implementing high-quality FPM reconstruction.