A trajectory simplification method based on the clustering of vessel speed and course segments

Abstract

Automatic Identification Systems (AIS) can provide massive ship trajectory data that is valuable for mining information in water traffic. However, it is also easy to cause the data redundancy, adding difficulty to the data processing. To address this problem, this paper proposes a trajectory simplification method based on the segmental clustering of the ship's speed and course. Firstly, the maneuver points of a trajectory are determined by the speed and course threshold. The trajectory is divided into segments in accordance with the maneuver points. Then, the feature points of each segment are extracted based on K-medoids clustering method. The maneuver points and feature points of every trajectory segments are combined as trajectory features s to achieve the purpose of simplifying the ship's trajectory. The experiment results show that the method can well maintain the feature of original trajectory and eliminate the redundant track points in AIS trajectory. Compared with tradition Douglas-Peuker (DP) algorithm, the distortion rate could be reduced 0.2% with the same simplification.