Fourier Ptychographic Microscopy Based On Optimized Illumination

When observing samples using a traditional microscope,the low magnification objective lens provides a wide field of view but low resolution,while the high magnification objective lens offers high resolution but a small field of view.Therefore,the challenge of achieving large field of view and high resolution simultaneously has been a difficult problem to solve.Fourier ptychographic microscopy(FPM)has become an important method to address this issue.It illuminates the sample with a programmable LED array,and under different angles of illumination,moves the spectrum information beyond the cutoff frequency into the aperture of the low magnification objective lens.After capturing a series of low-resolution images,the high-resolution and large field-of-view image can be reconstructed using a phase recovery algorithm.However,FPM requires capturing a large number of low-resolution images,which takes a long time,and this is clearly a weakness of the method.This paper mainly focuses on the research and exploration of LED illumination modes,and improves the imaging efficiency of FPM through the improvement of the illumination mode.Firstly,this paper elaborates on the basic principles of FPM,such as the principle of highresolution imaging,the optical path of imaging,and the process of reconstruction algorithm.At the same time,several imaging parameters and multiplexing methods of FPM are introduced.This paper addresses the issue of redundant spectrum information in FPM and proposes a novel FPM method that adaptively selects LEDs based on the image.The feasibility of this method is demonstrated through experiments.By analyzing the high-resolution and low-resolution images of the same sample,and simulating the FPM imaging process,225 secondary low-resolution images are generated using the high-resolution and low-resolution images as reference images.Their SSIM values are computed and analyzed,and it is found that the high-resolution and low-resolution images have similar spectral distributions.Based on this finding,the image captured by the LED in the center can be used as prior information to predict and analyze the spectrum of the reconstructed image,and important LED positions can be selected,greatly reducing the number of images required for FPM and improving its imaging efficiency.In order to obtain a better imaging quality and further accelerate the reconstruction speed,this paper proposes a multiplexing method based on optimizing the illumination pattern.Firstly,this paper uses an image quality evaluation method to obtain the difference matrix between FPM reconstructed images with and without illumination from a certain LED.Secondly,the distribution of important LEDs is analyzed using the pseudo-color map of the difference matrix,and an optimized illumination pattern is obtained that only illuminates the LEDs in the center area and the edge area.Finally,the multiplexing method is applied to the center area of the optimized illumination pattern to accelerate the reconstruction efficiency,and satisfactory reconstructed images are obtained.