A maritime traffic route extraction method based on density-based spatial clustering of applications with noise for multi-dimensional data

Maritime traffic route extraction is essential for analyzing dynamic ship navigational information in vast sea areas. However, maritime traffic route extraction is still challenging due to the high frequency and freedom of ship navigation. Thus, our study proposes a maritime traffic route extraction method based on automatic identification system (AIS) data and the multi-dimensional density-based spatial clustering of applications with noise (MD-DBSCAN) data. In the first part of the method, the Douglas–Peucker (DP) algorithm is used to compress massive ship trajectories, and a new indicator (the average compression score (ACS)) is used to determine the optimal compression threshold. In the second part, with the measure of similarity between ship trajectories by two similarity measure indicators, including the spatial distance and the difference of the course over ground (COG) of the key turning points of the ship trajectories, the ship trajectories are clustered by the MD-DBSCAN algorithm. In the third part, the centerlines of each cluster are extracted with the triangular network center method. Numerical experiments are conducted based on massive AIS data in the South China Sea. The experimental results show that our method dramatically influences the recognition of noise and normal trajectories and can effectively extract maritime traffic routes.