Single-Photon Reflectivity And Depth Imaging By Continuous Measurement Of A Rrival Time Of Photons

A single-photon counting Lidar comprised a high repetition rate pulsed laser and a single-photon detector with single photon sensitivity.The reflection intensity and depth of the surface of the long-distance object can be recovered by counting the number of photons in the laser echo signals and measuring time-of-flight(TOF)of the photons.Time-correlated single-photon counting(TCSPC)is an extremely weak light signals detection technology.The combination of the single-photon counting Lidar system and TCSPC technology can greatly improve the detection sensitivity,allowing the use of lower power semiconductor lasers for long-range detection.It is especially in the detection of long-range objects in the atmospheric environment,and airborne remote sensing,etc.have important applications in scenes with strict limits on weight,size and volume."We combine the single-photon counting Lidar system and TCSPC technology to design a single-photon reflectivity and depth imaging by continuous measurement of arrival time of photons.The main research contents and results are as follows:(1)A single-photon reflectivity and depth imaging by continuous measurement of arrival time of photons is built.A photons' arrival time measurement method that continuously measures at a scanning position and has a common starting point is proposed.The number of laser pulses is counted by a specially designed field programmable gate array(FPGA)control module as the coarse time of arrival photon.Time interval between arrival photon and the nearest coming laser pulse is measured by a time-correlated single-photon counting(TCSPC)module as the fine time of arrival photon.Using the system,not only the single-photon counting imaging can be realized,but also the first photon imaging,the first two photons imaging,the first three photons imaging,etc.can be realized.(2)A model of reflectivity and depth imaging based on the doubly stochastic Poisson point processes is established and a Monte Carlo simulation is performed.The photon statistical model based on the doubly stochastic Poisson point processes,the time-gated filtering algorithm and the reflectivity algorithm based on maximum likelihood estimation are derived.The Monte Carlo method is used to simulate the photon statistical model,the time-gated filtering algorithm and the objects reflectivity and depth images reconstruction algorithm,as well as the influence of light intensity,noise level and scanning time on the imaging performance.(3)We have carried out an experimental study on the calibration of reflectivity and depth imaging performance,and conducted reflectivity and depth imaging experiments on objects in different types of scenes.The system achieves high-sensitivity imaging and achieves a pixel resolution of up to 512×512 pixels for reflectivity and depth imaging.The experimental results show that the horizontal spatial resolution is 2mm,the vertical depth resolution is 5.375 cm,and the average number of photons per pixel is less than 1.3 photons.(4)A fast pixelwise non-local means Poisson denoising algorithm based on iterative variance stabilizing transformation(IVST-FPNLM)is proposed.Combining variance stabilizing transformation and fast pixelwise non-local means denoising algorithm,the Poisson denoising experiments are performed on the reflectivity images that obtained by the system in the,and compares the denoising performance with the classic Poisson denoising algorithm.The experimental results show that the IVST-FPNLM algorithm has better denoising effect on subjective evaluation and objective evaluation index.