Application Research Of Off-axis Digital Holographic Microscope 3D Quantitative Imaging

Off-axis digital holographic microscopy is a non-invasive,real-time high-resolution quantitative phase imaging technique that enables physiological and pathological activity monitoring of living cellular tissues and microstructured surfaces by digitally recording and reconstructing the amplitude and phase components of the wavefront Profile measurement.Based on the existing theory of digital holographic microscopy,this thesis mainly conducts in-depth research on expanding the imaging field of view,improving the stability of the recording optical path,improving the quality of the hologram,and improving the speed of the phase reconstruction algorithm.The main contents include:1.Aiming at the problems of poor real-time performance and low reconstruction accuracy of commonly used phase reconstruction algorithms,a phase unwrapping method based on the transmission equation of light intensity was proposed.The simulation and experimental results show that the unwrapping algorithm of the light intensity transmission equation solved by discrete cosine is more efficient than the traditional method without losing the accuracy.And the method shows stronger robustness for dealing with objects with sharp changes in phase.2.Aiming at the problems of poor stability and inconsistency between the field of view and high resolution of the traditional DHM system,a dual-wavelength commonpath off-axis digital holographic microscopy quantitative phase imaging method is proposed.Acquire two phase images at different wavelengths.This method further expands the field of view of the original camera by nearly 74.0% while keeping the imaging resolution of the original system unchanged,which is more compact and simpler than the previous method using field of view flipping;The common channel structure and the method of multi-wavelength multiplexing bring higher stability to the system.Finally,the feasibility and effectiveness of the proposed method are proved by experiments.3.Combining off-axis DHM technology and deep learning,first of all,for the zeroorder noise interference problem of traditional off-axis DHM,this these proposes a zeroorder image suppression method for digital holographic microscopy based on Pix2 pix generative adversarial network,and introduces data generation The method and the generalization ability of the generative network are tested.The experimental results show that the method can effectively suppress the autocorrelation term(ie zero-order image)of off-axis holography.Furthermore,the network exhibits excellent robustness to holograms with different degrees of background noise and phase distortion;Finally,a digital holographic red blood cell micrograph phase reconstruction method based on U-net network is proposed,the technical advantages and future challenges of this method are analyzed and discussed,and the experimental results show that the network can be used to achieve The structural characteristics of blood cells and the learning of optical path quantification information are obtained,which endows ordinary optical microscopes with quantitative phase imaging capabilities,and at the same time reduces the influence of speckle noise caused by high-coherence light sources to a certain extent.