Two-Parameter Gauss-Newton Based Real- Time Ranging Method for Full-Waveform LiDAR

Real-time ranging can boost the imaging performance and speed for full-waveform light detection and ranging (FW-LiDAR). To enhance ranging accuracy and maximum ranging rate, a two-parameter Gauss-Newton (Two-GN) based real-time ranging method is proposed for high- accuracy real-time ranging with lower hardware resource consumption. The Two-GN reduces the coefficient matrix size from 3×3 to 2×2 for obtaining quick iteration step vector solutions. FPGA implementation employs a module-level pipeline to enhance the maximum ranging rate. Experimental results showed that the ranging standard deviation (RStD) of Two-GN achieves 1.5 mm@51.90 dB, the absolute of mean ranging error (MRE) is lower than 2.0 mm@12.53 dB. The maximum ranging rate is improved to 2.21 MHz, the FPGA hardware resource and power consumption are reduced by over 30%. It is beneficial for real time imaging applications in autonomous driving and spacecraft space docking.