A Fast Close-Target Ranging Method for LiDAR in Fog Using Gauss–Newton Global Optimization

Light detection and ranging (LiDAR) is negatively affected by target signal attenuation and fog clutter interference in foggy conditions, complicating the accurate target extraction from fog-and-target overlapping waveforms. To address the above issues, a fast close-target ranging method based on Gauss-Newton global optimization is developed to extract additional target points from fog-and-target overlapping waveforms. The proposed method innovatively employs adaptive selection of the optimal initial value, global optimization for overlapping waveform fitting, and Gaussian-Newton (GN) iteration for fast convergence, all of which significantly improve target detection rates, ranging accuracy, and processing speed. Validation experiments are conducted in a long-and-controllable fog chamber using a lab-developed full-waveform LiDAR system. Results demonstrate that the proposed method achieves a target extraction rate above 93%, a ranging accuracy and precision within 1.5 cm, and a run time of 0.6 ms per waveform in heavy foggy conditions (visibility between 8 and 32 m), outperforming existing techniques. The presented method is well suited for fast processing and real-time interference-resistant LiDAR imaging in adverse weather, potentially broadening the application of LiDAR to diverse harsh environments.