Abstract: Objectives To address issues such as insufficient positioning accuracy and limited perception capabilities during ship berthing and unberthing operations, this paper proposes a high-precision berthing status perception method based on shipborne LiDAR. Methods First, a unified coordinate system and time synchronization strategy are established to perform filtering and preprocessing of point cloud data. Second, a point cloud registration method based on a combination of curvature and principal component analysis for feature extraction is proposed, and a constraint model of point-line and point-plane relationships is established to achieve pose estimation. Finally, the berth front is extracted using a random sampling consistency algorithm based on visual prior constraints, and pose compensation is introduced to calculate the berthing distance and berthing angle. Results Experimental validation was conducted using data from the sea trials of the “Xin Hongzhuan” vessel. The results show that the algorithm achieved an average absolute trajectory error of 0.274 m and an average attitude angle estimation error of less than 0.29°; the average berthing distance error between the vessel and the berth was less than 0.032 m, and the average berthing angle error was less than 0.25°. Conclusions This demonstrates the accuracy and robustness of the berthing state detection algorithm, providing reliable support for autonomous berthing and unberthing control in smart ships.