Full Digital Lidar Waveform Processing And Match

Li DAR is an proactive detection technology that can quickly acquire threedimensional information of earth surface and ground targets.Frequency,range and accuracy are three important indicators for measuring Li DAR's capability.With the applications of Li DAR in national mapping,industrial measurement and relics restoration,people have put forward higher requirements for the measurement efficiency,distance and laser foot density.The traditional Li DAR is limited by the echo receiving range and the measuring frequency,resulting in narrow laser measuring range,small measurement density and low measurement efficiency.The full-digital Li DAR can fully digitizes all backscattered echoes in the laser foot and provides richer details.This paper is based on the problems encountered in the participated projects,including: the dynamic receiving of echo signals,the processing and registration of waveform data and ranging ambiguity in the high-repetition and long-range Li DAR.The main research contents and innovations include:1.Study the waveform processing flow in full digital Li DAR,including waveform noise reduction,waveform detection,waveform fitting and waveform decomposition.Used the low-pass filtering method to complete the waveform denoising,studied adaptive threshold detection and multiple waveform modeling methods.and use LM and EM methods to complete the waveform decomposition;2.For the problem of wide dynamic signal reception in full digital Li DAR,a two-way splitting and multi-stage amplification method is proposed and corresponding waveform storage structure is designed to solve the problem of wide range signal reception and complete high efficient waveform data storage;3.For the high precision ranging problem in full digital Li DAR,interpolation filtering is used to complete the up-sampling of the waveform data,realizing high-precision time registration,and the correction model of the amplitude-phase errors with waveform echo characteristics was established to improve the measurement accuracy;4.Aiming at the problem of ranging ambiguity in high transmitting frequency Lidar,a method combining neural network and binary pseudo-random code was proposed.The accurate corresponding relationship between the seed pulse and the echo is calculated to decide target distance.Through the research and experiments above,the real laser radar waveform data is processed and analyzed.The optical design with two-way splitting and multi-stage amplification extend the echo reception range to millions level,and the single waveform data storage is only 12% of the original.Under the condition of 1.2 GHz sampling,the accuracy of the method combined up-sampling and the registration waveform is up to mm level and the method of pseudo-random modulation solves the problem of ranging ambiguity in high-frequency and long-range measuring.