The Research Of Fourier Ptychography Algorithm Based On Image Sparse Representation

Fourier ptychography is a new imaging technique that integrates the theory of structured illumination,ptychography and phase retrieval,etc.The Fourier ptychographic microscopy invented by this technology can achive high-resolution and large field-of-view imaging with low numerical aperture and low-resolution microscope objective.However,the traditional Fourier ptychography algorithm collects data for longer acquisition time and the anti-noise performance is poor.How to improve the quality of reconstructed images and anti-noise performance while shortening the acquisition time is the focus of the current research on Fourier ptychography algorithm.Based on the theory of compressed sensing and phase retrieval,this paper studies the Fourier ptychography algorithm utilizing sparse prior knowledge.In this paper,the main research contents are as follows:Firstly,aiming at the information redundancy and preserving invariance of undecimated wavelet transform,the undecimated wavelet transform is used as the sparse prior.This paper proposes a Fourier ptychography algorithm based on undecimated wavelet regularization and utilizes the alternating direction multiplier method to solve the optimization problem.The experimental results show that this algorithm can effectively improve the anti-noise performance and quality of the reconstructed images compared with the traditional Fourier ptychography algorithm.Secondly,due to the approximate shift invariance and multi-directional selectivity of the dual-tree complex wavelet transform,this paper combines the dual-tree complex wavelet regularization with the traditional Fourier ptychography algorithm and proposes a Fourier ptychography algorithm based on dual-tree complex wavelet regularization.The alternating direction multiplier method and gradient descent method was applied to solve the optimization problem.In the experiment,the effectiveness of the algorithm is verified by reconstructing multiple images.Finally,the total variation has good characteristics of preserving the edge of the image and suppressing the oscillation effect.Therefore,this paper employs the sparse prior of images under total variation and proposes a Fourier ptychography algorithm based on the total variation regularization.Due to the defects of more variables of the alternating direction multiplier method,the gradient descent method is exploited to solve the optimization problem.The experimental results show that the proposed algorithm is effective.