The last decade has witnessed the prevalence of sensor and GPS technologies that produce a sheer volume of trajectory data implying the city dynamics and human behaviors. Detecting trajectory anomaly is undoubtedly one of the most important tasks in trajectory data management since it serves as the foundation of many advanced analyses such as suspicious behaviors identification during the important events. Tremendous efforts have been spent on this topic, however previous works always require time-consuming preprocessing or are susceptible to the model parameters. In this paper, we first measure the similarity between trajectories through dynamic time warping and then resort a robust density based clustering method, ordering points to identify the clustering structure, to distinguish the anomaly cases. The real-world taxi trajectory dataset in Beijing validate the effectiveness of our approach.

CICTP