Waveform centroid discrimination of pulsed Lidar by combining EMD and intensity weighted method under low SNR conditions

In this study, we report on a self-adaptive waveform centroid algorithm combing EMD and intensity-weighted (IW) method for the accurate Lidar ranging under low signal-to-noise ratio (SNR) conditions. Empirical mode decomposition (EMD), discrete cosine transformation (DCT) and inverse discrete cosine transformation (IDCT) are utilized to filter out the high-frequency interference in Lidar pulse signal. A time window is set to adaptive select the effective data. The centroid points are calculated by the intensity-weighted (IW) waveform centroid algorithm with the selected data. The proposed algorithm is experimentally tested, achieving an average error of about 153 ps, 209 ps and 232 ps under the SNR of 5.1 dB, 2.7 dB and 1.8 dB, respectively, which exerts better precision compared to the wavelet soft thresholding with intensity-weighted (WL-IW) algorithm and EMD with peak selection (EMD-PK) algorithm. Furthermore, the proposed algorithm is fairly robust with remarkable detection success rates of above 96% under low SNR conditions. The proposed waveform centroid algorithm has the potential for pulsed Lidar's ranging tasks in harsh environment.