Lidar full-waveform decomposition based on empirical mode decomposition and local-Levenberg-Marquard fitting

The light detection and ranging (LIDAR) full-waveform echo decomposition method based on empirical mode decomposition (EMD) and the local-Levenberg-Marquard (LM) algorithm is proposed in this paper. The proposed method can decompose the full-waveform echo into a series of components, each of which can be assumed as essentially Gaussian. The original full-waveform echo is decomposed into the intrinsic mode functions (IMFs) and a final residual by using the EMD first. Then, the average period (T_m¯) and corresponding energy densities (EDs) of all IMFs are calculated. A suitable IMF is selected based on the relationship between the EDs of IMFs and the white-noise theoretical spread lines of the 99% confidence-limit level. The components in the full-waveform echo can be detected according to the positions of the maxima of the selected IMF. The initial parameters are estimated by using local-LM fitting. The initial parameters are fitted by global-LM fitting. Compared to the traditional (zero-crossing) ZC method, the proposed method has strong anti-noise performance. It can precisely detect the components and estimate the initial parameters of the components. The proposed method is verified by using the synthetic data; coding LIDAR recorded data; and Land, Vegetation, and Ice Sensor data.