Simplification method of ship circular trajectory based on classification strategy

The Automatic Identification System (AIS) offers an extensive database for maritime research. However, the raw AIS data often contains redundancies. Utilizing this data directly leads to inefficient use of storage space and higher computational demands, necessitating the implementation of data compression. In this paper, we propose a simplification method for ship circular trajectories based on a classification strategy to improve the accuracy of trajectory compression and the ability to retain important feature points. Aiming at the problems that the Douglas-Peucker (DP) algorithm may miss the port feature points and insufficient retention of the feature points of multi-turn paths when dealing with circular trajectories, this method firstly classifies the trajectories initially according to the ship's speed, and then further classifies them finely according to the ship's orientation, and the fitness function is introduced in the compression. Through this classification and compression strategy, port characterization points and turning feature points can be effectively identified and retained. Experimental results on real datasets show that this method outperforms traditional methods in terms of average azimuth error and compression accuracy and is able to better retain important trajectory features.