Study On Time Domain Statistical Focus Evaluation For Red Blood Cell Digital Hologram Reconstruction

Blood cells have the functions of transporting oxygen,defending against germs,clotting,etc,and the detection on them can provide the useful information for disease prevention,diagnosis and treatment.Optical microscope has become the most widely used blood cell detection method due to its advantages in visual detection.However,its complicated manual operation,small imaging field of view,complicated hardware system,difficulty in integration,and the inability to achieve three-dimensional imaging have limited its development.Lensless digital holographic imaging technology uses an image sensor to record the interferometric pattern of objects,and restores the true three-dimensional original image on the computer.It can circumvent the shortcomings of optical microscopy in blood cell imaging analysis and has found wide applications in this field.The calculation of the parameters of blood cells and the morphology observation are based on the digital holographic reconstruction image,and the quality of the reconstruction image is closely related to the reconstruction distance.Only when the reconstruction distance is equal to the object distance during hologram recording can an accurate object image be obtained.However,it is usually difficult to measure the recording distance directly.Thus automatic focusing is required to determine the best image plane.Therefore,systematic research on the focusing process and the related parameters is one of the important research directions in the field of lensless digital holography.This paper focuses on the reconstruction of the red blood cell digital hologram,and conducts systematic research on the focus search algorithm,focus evaluation function,and focus search parameter determination involved in automatic focusing.First of all,this paper introduces the basic principles of three focus search algorithms: Traversal search algorithm,Fibonacci search algorithm and “blind” climbing search algorithm.Through analysis,the traversal search algorithm is most suitable for research in this article;Then it focuses on the evaluation basis of the existing focus evaluation functions,classifies these functions according to different evaluation principles,and introduces the most widely used time domain statistical focus evaluation function.Then Using the self-built lensless digital holographic imaging system to obtain red blood cell holograms,and then using matlab to program algorithms and discuss the effect of focus search parameters on the focus process,including three reconstruction algorithms of Fresnel reconstruction,Convolution reconstruction and Angular Spectrum reconstruction,the size and position of the focus image,the influence of search range and search step.Through discussion,the optimal focus search parameters are determined and applied to the autofocus test of the time-domain statistical focus evaluation function,and then use qualitative and quantitative evaluation indicators to make a comprehensive comparison of its focusing performance.The experimental results show that Angular Spectrum reconstruction performs better than Convolution reconstruction and Fresnel reconstruction in focus search.The size of the representative focus area are more appropriately selected as 256×256 pixels,and the position of the focus search area does not affect the focus result,so you can choose any position to focus.The search step is determined to be 0.01 mm.For the search range,1.1～2.1mm can be selected in practical applications.If it is used to analyze and compare the performance of the focus evaluation function,0.6～2.6mm is more suitable.In addition,different functions have different focusing performances in holograms of different qualities,but SI is the best time-domain statistical focusing evaluation function.For holograms containing impurities,impurities’ information will overwrite cells’ information,resulting in the failure of all focus evaluation functions.The study also found that the overall noise immunity of the time-domain statistical focus evaluation functions is weak.In summary,this paper determines the optimal search parameters in the red blood cell hologram reconstruction focusing,and avoids the improper selection of the focus parameters to affect the focusing process,meanwhile,selects the best performance evaluation function under the optimal search parameters.This study also improves the focusing accuracy and the quality of the reproduced image,which is conducive to the development of related researches such as blood cell count,cell internal parameter calculation and morphological observation.