Study On Key Problems Of Macroscopic Fourier Ptychography Imaging

At present,when the imaging system pursues higher spatial resolution,it encounters the bottleneck of too large volume and weight,too large power consumption,complex structure and high cost.Therefore,improving the spatial resolution of the system is the long-term goal of imaging science.Fourier Ptychography Microscopy(FPM)is a new generation of computational imaging method proposed in recent years.This technology skillfully combines synthetic aperture and phase recovery technology,which makes it possible for coherent synthetic aperture imaging in visible light.The purpose of this dissertation is to explore Fourier ptychography systems that are more suitable for macroscopic imaging,and analyze its influencing factors and error correction methods.The main work of the dissertation is summarized as follows:(1)A Fourier ptychography(FP)imaging model based on laser variable-angle illumination is proposed.Compared with the traditional LED array,this method has the advantages of uniform light intensity,adjustable angle and longer lighting distance.An experimental platform based on the model was built to verify the 4 times spatial resolution improvement of the device in the 2.45 m transmission imaging mode.And based on the collimator,the 5 times spatial resolution improvement in the infinite-distance reflection imaging mode is verified.(2)Aiming at the registration problem of speckle noise images in the camera-scanning FP system,an algorithm with coarse registration based on correlation function and fine registration based on speckle fractal dimension is proposed.A set of reconstructed images of experimental data registered by this method can increase the resolution from 44.2 um to 35.1 um.(3)For the speckle noise caused by rough samples,a FP reconstruction algorithm based on the enhanced Lee filter algorithm is proposed.The algorithm uses the enhanced Lee filtering process as the outer loop of the phase retrieval iteration,which filters out speckle noise while retaining enough effective information.A set of experimental data is used to verify the effectiveness of the proposed algorithm.(4)Aiming at the spatially varying aberrations,a FP imaging method based on the phase diversity is proposed: The pupil function of the camera-scanning FP system is constantly changing for each acquisition process,and the traditional EPRY algorithm cannot effectively restore the aberrations.A FP imaging method based on the phase diversity is proposed,which can accurately reconstruct the aberration function corresponding to each position,thereby improving the imaging quality of the system.Moreover,this method is also very advantageous for future applications of the system in scenes with time-varying aberrations such as turbulence.A camera-scanning FP experiment platform was built,which verified 2.8 times spatial resolution improvement at 1.3 m imaging distance,and verified 2.3 times spatial resolution improvement at infinite imaging distance based on collimator.(5)The influence of the frequency of the illumination pattern and the number of pattern scanning on the pattern-illumination FP imaging are discussed.The higher cut-off frequency and more scan times of the illumination system will produce higher-resolution reconstructed images.A pattern-illumination FP experimental platform based on random spot is built,and the imaging performance of the device for smooth and rough samples is demonstrated.