Technology Research Of Three-dimensional Quantitative Phase Microscopy Imaging

In the study of living biological cells,the three-dimensional structure imaging technology often needs to be reflected by measuring the refractive index of cells,which is also a good method for imaging without contrast agent.In addition,for some fields,the refractive index of cells can also be used to analyze the cell state.Will stay with the method of the interference measuring cell interference imaging under the multiple perspectives,and get the phase information by the every location,integrating the information reconstruction,again at this time to get the required cell refractive index distribution,especially in cell microflow,it has been used quite a lot because the advantage of three-dimensional refractive index,so the three dimensional quantitative phase microscope could be extensively used.This paper based on the standard of the Mach-Zehnder holographic interferometer for optical fiber three-dimensional imaging quantitative phase microscope,using step motor to the rotation of the optical fiber samples,different location of the optical fiber is obtained phase,we in the heart of the program step motor and CCD,make each stepper motor rotation Angle CCD for filming in time,after such reduced due to different angles and the CCD shooting time caused by inconsistent error,a group of interference figure that good result is obtained.Will this group interference figure solution with the phase information of the fiber,the information of three-dimensional refractive index distribution of the fiber will be used to reconstruct,in this paper,we put forward a method of in-perot interferometer method of multiple beam interference streaks,and has carried on the simulation and experimental validation of the method,put in the Yu Fabu-the object under test perot cavity of multiple beam interference.Then,the Fabry-Perot interference fringes obtained from different angles were used to reconstruct the three-dimensional refractive index distribution of the cells to be measured by the combined system of stepping motor and CCD.The results show that the interferometric method can reduce the error of phase reconstruction and improve the accuracy of 3D refractive index distribution.Later,this paper introduces the specific applications of 3D refractive index reconstruction and cell image transmission.Using the learning model obtained by deep neural network learning,coaxial dual-waveguide fiber and panda fiber are respectively applied to the learning model samples of residual neural network,and the experimental results of the two kinds of fiber are obtained.Compared with the traditional filter backprojection method,it is found that the learning model trained by the deep neural network is obviously better than the result obtained by the filter backprojection method in the optimization of optical fiber three-dimensional refractive index distribution,and it will play an important role in medicine and biology.