Screening Of Laser Footprints Based On The Full-waveform Gaussian Decomposition

High-precision laser ranging value is the key and guarantee for the realization · of large scale photogrammetry mapping without ground control point. The high-precision laser ranging is used as the control constraint condition to participate in the block adjustment calculation,and then the ground elevation accuracy can be improved. When the relief is big or geographical objects are very complex in laser footprint, the range measurement and laser footprint positioning accuracy will be affected. In the paper, gaussian decomposition of full-waveform LiDAR data is carried out by using the optimized LM (Levenberg-Marquard)algorithm. On this foundation, the physical characteristics of target in laser footprint are obtained. Then screening criteria are established by making screening experiments according to the requirement of 1:10000 mapping accuracy, and thus obtaining the laser footprints which meet the requirements of large scale mapping without ground control point.Therefore, the screening of laser footprints has an important role in study of enhancing elevation accuracy in the succeeding work.The main contents are as follows:(1)Airborne LVIS and onboard GLAS data format are described in detail and files are parsed according to the format. This is the basis of the topic research.(2) Using the threshold method and gauss filtering comprehensively to remove noise from signal and experimental results show that the proposed method can effectively eliminate noise while preserving the effective signal which provided data support for the subsequent waveform decomposition.(3)The current research status of full-waveform gaussian decomposition and the principle of traditional LM algorithm are analyzed. And optimizing the LM algorithm by means of updating the Jacobian matrix or not, and adjusting the damping coefficient. Then make experiments by using the traditional LM and optimized LIVM algorithm with LVIS and GLAS data. The experimental results show that using the optimized LM algorithm can simplify the calculation process and the simulated waveform can be fit well with the original waveform.Then assist screening the laser footprints in the follow-up study.(4)Study on the screening of laser footprints to meet the precision requirement of 1:10000 photogrammetry mapping without ground control point. Simulation analysis on the influence of terrain and surface features on echo signal, and the physical characteristics of target in laser footprint are obtained through the full-waveform gaussian decomposition,and study on the influence of slope on ranging accuracy. On this foundation, screening criteria are established by making screening.experiments. The experiment is carried out by using 30 GLAS footprints and Google earth image in Wendeng District Weihai. Finally the laser footprints are obtained which meet the requirements of large scale mapping without ground control point.