Research On Aberration Correction Algorithms And Techniques In Holographic Microscopic Imaging

Digital holographic microscopy is a non-destructive,label-free,accurate and near real-time interferometric technique for acquiring quantitative phase images.Thus far,it has been used in applications such as cell biology analysis,microstructural measurement,particle tracking,microfluidics metrology,and neural science.To improve the spatial resolution of digital holographic microscopy measurements,microscope objective lenses are inevitably used for image magnification.However,the misalignment of the microscope objective and the defects of all optical components,including the microscope objective,can lead to aberrations in the measurement results.Therefore,aberration compensation and correction are an indispensable part of digital holographic microscopy.Researchers have proposed many methods for aberration compensation and correction,which can be divided into physical methods and numerical methods based on the way they are implemented.However,the current aberration compensation methods always have their own limitations and shortcomings.For example,methods that require multiple measurements,such as the double exposure method,have higher requirements on the measurement environment and system stability,that is,the robustness of the method is lower;the typical method based on Zernike polynomial fitting has the influence of the loss of polynomial orthogonality which leads to the fitting error when fitting irregular aperture data;methods that requires the separation of the sample and the background needs to use the phase of the background area to correct the aberration,which limits the measurement object to only the sample with a relatively simple shape and a small space in the field of view;methods based on polynomial fitting,parameter optimization or iteration require high computation time and cannot be applied to applications with high real-time requirements.In short,the current aberration compensation methods mainly have the problems of low robustness,irregular aperture fitting error,difficulty in correcting and compensating full-aperture complex samples,and time-consuming compensation methods.Therefore,it is necessary to select the most suitable aberration correction method or propose a new aberration correction method according to the different measurement methods and measurement targets.Based on the current situation that there is no general aberration correction method,this thesis focuses on the aberration correction method in digital holographic microscopy,and explore the aberration correction methods under different measurement situations and measurement requirements,such as high robustness and rapid correction,high-precision correction of irregular apertures,full-aperture complex sample correction,and real-time correction.In these works,we proposed some aberration correction methods with higher applicability,fill some gaps in the current aberration correction methods.systematically solve the impact of background aberrations on digital holographic microscopy imaging measurements,and provide the new theoretical reference for the selection of aberration correction methods in practical applications.The main innovations and research results of this thesis are as follows:(1)This thesis proposed an off-axis holographic aberration compensation method based on self-extension of holograms:We remove the fringes of the sample area in the hologram by the background recognition method based on corner detection,then expand the fringe to fill the removed area of hologram by self-extension based on Fourier transform.That is,the sample-free reference hologram which contain the total aberration of the system is obtained,and the phase aberration compensation is achieved by subtracting the system background aberration from the measured phase containing the aberrations.The final aberration compensation of this method is the same as the double exposure method,both of which realize the aberration compensation through a background hologram.However,the method proposed in this thesis obtains the background hologram by numerical calculation,and does not require the second hologram acquisition,which avoids the influence of errors that may be introduced by factors such as light source jitter,optical component displacement or vibration in the system in the interval between two acquisitions.Therefore,the method proposed in this thesis has high measurement robustness,but is only suitable for the measurement of samples with simple shapes and clear backgrounds due to its requirement for background recognition and the ability of current background recognition algorithms:(2)This thesis proposed an automatic high order aberrations correction for digital holographic microscopy based on orthonormal polynomials fitting over irregular shaped aperture:We obtain the aperture shape of the background phase data required for the current fit by a corner detection-based background recognition method,and derive the specific orthogonal polynomials that still maintain orthogonality under irregular target apertures.The expansion coefficient obtained based on the fitting of this polynomial is converted into the expansion coefficient of the Zernike polynomial under the unit circular aperture through the coefficient transformation matrix,so as to calculate the aberration mask for aberration compensation.By deriving a specific orthogonal polynomial to fit the aberration,the fitting error caused by the lack of orthogonality of the traditional unit circle aperture Zernike polynomial under the special-shaped aperture is avoided,so that the phase aberrations of low-order and high-order terms under irregular data apertures can be well compensated.The method breaks through the traditional polynomial fitting-based method’s requirement for data aperture shape,can automatically derive polynomials that maintain orthogonality and implement aberration fitting according to the background data aperture shape that needs to be fitted in current actual measurements,and improves the accuracy of aberration compensation in actual measurements,and can be used for measurements with irregular data apertures;(3)This thesis proposed a sequential shift absolute phase aberration calibration in digital holographic phase imaging based on Chebyshev polynomials fitting:We introduce the basic idea of absolute detection method in traditional optical surface inspection into phase aberration compensation of digital holographic microscopy.The phase and system aberrations of the measured object are separated by two specific displacements of the sample.Then,based on the relationship between the aberrations measured at different positions,the phase of the aberration is calculated by the Chebyshev polynomial fitting to realize the compensation of the aberration.This method breaks through the dependence of the traditional polynomial fitting-based method on the background recognition algorithm and the limitation that the proportion of the tested sample in the field of view must be small.And it uses all the data in the field of view instead of only fitting the background phase data,which preserves the integrity of the data,that is,effectively protects the mid-and high-frequency details of the sample under test in the phase image.Therefore,this method can be applied to complex and large-area sample measurement,even if the sample fills the entire detection field of view;(4)This thesis proposed a method based on sequential shift and a differential-integral algorithm to compensate for fast phase aberrations:The measured sample phase and system aberration are separated by 1-2 sample displacements.Then,the phase derivative of the tested sample in the corresponding displacement direction is obtained by simple subtraction,and finally the extraction and reconstruction of the true phase of the tested sample is realized by simple integration.The proposed method only uses simple difference and integration to process the data,avoiding the extreme computational cost problem always caused by complex and time-consuming polynomial fitting and optimization iterative methods,so that the true phase extraction of the measured object can be achieved at a near real-time speed.Therefore.this method is suitable for situations where real-time quantitative measurements are required.