Inshore-ship berthing parameters extraction system using 2D bird's eye view of 3D LiDAR data

The accurate and rapid estimation of berthing parameters is crucial for assisting the autonomous berthing of inshore ships. Recently, the 3D light detection and ranging (LiDAR) has proven highly effective in maritime monitoring, particularly for ship berthing, owing to its ability to provide detailed spatial position information of the target with high measurement accuracy. However, although raw point cloud contains rich features to perform information extraction, bird's eye view (BEV) projection, a more compact representation, is often preferred in order to meet the time and accuracy requirements and reduce computation costs in such real-time applications. Therefore, this paper presents a new method for estimating motion pose information of inshore ships using BEV image generated from LiDAR point cloud data. The method uses BEV images as input to determine the spatial position of ship's key points. And the real-time parameters, including the berthing angle, offshore distance, velocity and other information, are finally output by analyzing the differences in pose between these points across two consecutive frames. The proposed method is tested on the real-ship data collected at Dalian Port (China) and publicly available simulation data. Comprehensive experiments demonstrate that our method can provide more accurate and stable berthing pose information in real time, supporting efficient and safe berthing operations.