Lidar Full-Waveform Decomposition Based on the Empirical Mode Decomposition and Gaussian Function Model

Wu Qinqin,Qiang Shengzhi,Li Xicai,Wang Yuanqing

DOI: https://doi.org/10.1088/1361-6501/ab44e6

2019-01-01

Abstract:The lidar full-waveform echo can be decomposed into a series of components each of which can essentially be considered as Gaussian function models. In this paper, we use empirical mode decomposition to decompose the original full-waveform echo into a series of intrinsic mode functions (IMFs) and a final residual. The noise-level of the echo is estimated based on the first IMF. The IMF energy density is then compared to the white-noise theoretical spread lines of the 99% confidence-limit level to determine whether some IMFs are considered to contain information. A suitable IMF is then selected based on the comparison result. The maxima of the selected IMF can be used to predict the number and positions of the Gaussian components in a full-waveform echo. The other parameters can also be further estimated. The iteration decrement method is used to detect the components and estimate the initial parameters gradually. Simulations and experiments were done to verify the proposed method, and it was also verified by using the LVIS data set. The proposed method avoids having to use the zero-crossing of the first or second derivatives of the denoised waveform to estimate the initial parameters. The initial parameters are optimized by using the Levenberg?Marquardt algorithm.