Clustering Approach for Trajectory Anomaly Detection

The last decade has witnessed the prevalence of sensor and GPS technologies that produce an enormous volume of trajectory data describing city dynamics and human behaviors. Detecting trajectory anomalies 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 behavior identification during important events. Tremendous efforts have been spent on this topic, but previous studies have required time-consuming preprocessing or are susceptible to the model parameters. In this paper, we first measure the similarity between trajectories through dynamic time warping. We then use a robust density based clustering method of ordering points to identify the clustering structure, to distinguish the anomalous cases. We used a real-world taxi trajectory dataset in Beijing to validate the effectiveness of our approach.