Research On Three-dimensional Active Imaging With Polarization-modulated Method

Three-dimensional(3D)lidar system can produce a 3D representation of a scene’s 3D structure.Due to the characteristics of high resolution,rich information,fast imaging and all-day working,it has been playing an important role in target monitoring and recognition,large-sized objects’ precise measurement,non-contact & non-destructive detection,micro-vibration sensing and so on.However,there are still some shortcomings in the existing 3D imaging lidar systems:(1)For the 3D imaging lidar based on scanning mechanism,the imaging speed is limited by the scanning mechanism,which makes it hard to meet the need of dynamic imaging;(2)For the 3D imaging lidar based on avalanche photodiode(APD)array,the transverse resolution is limited by the detector and the corresponding readout integrated circuit(ROIC);(3)For the 3D imaging lidar based on intensified CCD(ICCD)camera,the transverse resolution and sensitivity is limited by the coupling procedure including photon-electron conversion,electron-photon conversion and photon-electron conversion again.Therefore,it has become a significant issue to provide an outstanding performance for the 3D imaging lidar system with high resolution,fast imaging,long ranging and low data rate.To perform outstanding performance for the 3D imaging lidar system,a polarization modulated 3D active imaging method is proposed in this paper.Firstly,with the polarization modulated characteristics of large aperture electro-optic modulators,a pulsed-laser ranging theory based on polarization modulation are established.Secondly,a simulation platform for the polarization modulated 3D imaging is established,with which both numerical analysis and simulation validation about the ranging theory are implemented.Finally,an experimental system for the polarization modulated 3D imaging is established,with which 3D images for static and dynamic targets at long-range are reconstructed and displayed,respectively.The main features and innovations for the 3D active imaging method in this paper are as follows:Firstly,a high-sensitivity and high-resolution 3D imaging method based on polarization modulation is proposed.With polarization-modulated technique,time information can be convert into gray information,which makes it possible for the high-sensitivity and high-resolution EMCCD cameras to perform 3D imaging.With the EMCCD cameras,the imaging procedure includes photon-electron conversion with one time,while that in the ICCD cameras includes multiple conversions due to the coupling structure.Such excellent feature makes the EMCCD cameras perform higher quantum efficiency and higher transverse resolution,especially suitable for high-resolution and long-range imaging.Based on dual EMCCDs structure,a 3D image can be reconstructed from the two modulated images,which make our framework act as a flash lidar system.Furthermore,an exposure time with sub-microsecond can be implemented with the polarization modulated technique,which will enhance the performance of dynamic imaging,regardless of objects motion or platform motion.Thirdly,an adaptive range-gated 3D imaging method is proposed in this paper.Based on the fact that range accuracy is inversely-proportional to the ranged-gate,the adaptive ranged-gate can be controlled to perform coarse range imaging or fine range imaging successively.Specifically,coarse range imaging is used to search for targets within a large range,while fine range imaging is used to improve the range accuracy by compressing the ranged-gate.Therefore,3D imaging can be achieved in higher range accuracy and wider gated-range.To reconstruct a 3D structure from two polarization-modulated images,an efficient sub-pixel image registration based on first principal component analysis(FPCA)is presented,with which a real-time 3D image can be reconstructed and displayed with higher efficiency and lower memory size,while maintaining the accuracy and robustness.In addition,a range noise suppression method based on the signal-to-noise ratio(SNR)of gray image is presented as well,which makes it possible to accurately identify 3D structure of a target from an obscure depth image.Such method can be implemented simply,which is also helpful for real-time 3D reconstruction and display.In summary,the proposed 3D imaging lidar system in this paper can provide higher resolution,faster imaging,longer ranging and lower data rate,which will outperform the existing system and open up a new way for 3D active imaging.