Research on anomaly trajectory detection method based on multi-feature clustering
With the rapid development of intelligent perception,wireless positioning and Internet information technology,the trajectory data of all kinds of moving objects has shown explosive growth. Due to the existence of multiple error factors,the trajectory data usually contains outliers,and the existing trajectory anomaly detection technologies usually ignore the local characteristics of the trajectory,which affects the accuracy of data analysis. This paper proposes an anomaly detection method based on multi-feature clustering,which combines the spatial,temporal,kinematic and statistical characteristics of the trajectory with the identification of stop points and the division of sub-trajectory,and uses adaptive parameter clustering to detect the local and global anomaly trajectory. The results show that this method can effectively identify abnormal trajectories,with high efficiency and accuracy,and has certain application potential in urban management,road planning,intelligent driving and other fields.
万方数据
维普
