Three-dimensional Imaging Via Time-correlated Single-photon Counting In Turbid Media

Imaging in turbid media is an important research direction in the current laser imaging field.The imaging method can image targets behind turbid media,which is of great significance for improving traffic safety and reconnaissance capabilities in harsh environments.In recent years,with the improvement of the sensitivity of detectors and imaging reconstruction algorithms,the requirements of imaging distance in turbid media are gradually increasing.And the demand for target recognition in turbid media is also increasing.However,the current traditional penetrating imaging methods have two problems.On the one hand,traditional detectors such as CCD and CMOS can only obtain the intensity information of the echo,and they cannot accurately obtain the temporal dimension information of the weak light field.On the other hand,traditional imaging methods only obtain target information from ballistic light,and removing the scattered light that is treated as interference.In this way,all the implicit target information is lost.This thesis constructs the transmission model of the light field in the turbid medium theoretically,and the transmission law of scattered light is realized through simulation.Then,the Time-correlated single photon counting(TCSPC)imaging system that can obtain the spatial-temporal dimensional information of the light field with high precision is constructed experimentally.Finally,this thesis studies how to use the algorithms to reconstruct the scattered light images to improve the quality of the ballistic light images from the algorithms.Firstly,this thesis presents a fast method to calculate the Mie scattering coefficients.This method can be applied to the Mie scattering process of light transmits in turbid medium.Then,based on this method,this thesis uses Monte Carlo to establish optical transmission model in turbid medium.With this simulation model,the spatial and temporal distributions of scattered light fields in turbid media are simulated.And it simulates the light field distribution at the receiving aperture including the scattering medium with different optical thickness and the targets with different spatial distance.It is also considered that the impulse response function and optical transfer matrix are essential in the algorithm recovery of blurred images.According to the simulation model,under the imaging scene of the scattering medium with specified optical thickness and constant size of photon packet statistical,the impulse response function and optical transfer matrix can be generated.Pave the way for the subsequent research of reconstruction algorithms.Secondly,this thesis adopts the TCSPC technology with the most potential application in the future to collect the echo signal of the targets in the turbid scattering medium.Therefore,this thesis builds two scanning TCSPC acquisition systems.One is the turntable scanning system and the other is galvanometer swing scanning system.Then,the two acquisition systems are used to obtain the target depth images through the liquid medium and gaseous medium under laboratory conditions.Results include the threedimensional depth images of single target and combined target.Comparing the experimental results with the simulation data to verify the reliability of the simulation model.Finally,according to the characteristics of the two scanning modes,giving adaptive detection environment respectively.Finally,this thesis introduces the reconstruction algorithms to restore the blurred depth images.First,the principle and recovery effect of the classic Wiener filter algorithm are analyzed,and by means of the backscattered light of the target,the image quality of this method is better than that of the traditional ballistic light imaging method.After that,this thesis tries to use GPSR and TV algorithms to restore the image,which proves they applicability.At last,we analyze the reasons of the unsatisfactory recovery effects by GPSR and TV algorithms,which provides a possibility for subsequent optimization of the recovery effects of these two algorithms.