Improved Gauss Inflection Point Matching Method for Lidar Echo Signal Decomposition

LiDAR is now widely used in measurement. In view of the current problem that the Gaussian decomposition method decomposes the full-waveform LiDAR superimposed signal incorrectly, this paper improves the Gaussian inflection point matching algorithm to decompose the LiDAR echo signal. First, it is necessary to find the inflection point corresponding to the echo signal. Secondly, the inflection point and peak value are used to match to find the main waveform. Finally, the iterative method is used to find the remaining parameters and bring them into the LM(Levenberg-Marquardt) algorithm for parameter optimization. By decomposing the echo signals in the two regions of Antarctica and the United States, the method in this paper decomposes the echo signals in the Antarctic region with a success rate of 99.9% and an average R2 of 0.9661. The success rate of decomposing echo signals in the United States is 88.7% and an average R2 is 0.8833.