Spatiotemporal Companion Pattern (STCP) Mining of Ships Based on Trajectory Features

Spatiotemporal companion pattern (STCP) mining is one of the means to identify and detect group behavioral activities. To spatiotemporal traveling ships from massive trajectory data understand movement law ships, this article proposes a feature-driven approach for STCP that consists (1) generating grid index via rasterizing geospace characterizing points sequences (STTGSs) ships; (2) designing filtering rules with constraints range, time distance construct candidate set ship mining; (3) measuring STTGS similarity associated setting confidence threshold realize mining. The effectiveness proposed method practically validated on real dataset which collected Taiwan Strait waters. experimental results are as follows: 825 pairs 225 accompanying mined when size 0.05° 0.5. Larger sizes can increase inclusiveness measurement, result in an pattern. A large number pseudo-accompaniment extracted set, resulting more dispersed distribution confidence. By verifying method, activities such cooperative operation, navigation so on, be detected. These provide reference research behavior identification have important application value water transportation management.