基于聚类的兴趣区域间异常轨迹并行检测算法

An Algorithm for Detecting Anomalous Trajectory Between Interest Regions Based on Clustering

轨迹异常检测能够用来分析移动对象的异常运动行为,在交通运输、医疗监护等领域都有广泛应用.兴趣区域是移动对象集中活动的区域.本文提出了一种新的兴趣区域间异常轨迹检测算法(Detecting Anomalous Trajectories Between Interest Regions,DATIR).不同于已有的从局部采样点进行检测的算法,DATIR算法综合考虑了轨迹的局部特征和全局特征,利用聚类方法检测兴趣区域间的异常轨迹,并能挖掘出兴趣区域间的正常路径.为了提高海量轨迹数据的异常检测效率,在DATIR算法的基础上,提出了一种并行检测算法(Parallel Algorithm for Detecting Anomalous Trajectories Between Interest Regions,PDATIR).实验结果表明,DATIR算法能够有效地检测兴趣区域间的异常轨迹,并且能够检测出兴趣区域间的正常轨迹;PDATIR算法在大数据集上表现出了明显的性能优势,具有较好的可扩展性和较高的加速比.

Trajectory anomaly detection can be used to analyze the anomalous behavior mobile objects,and it has been widely used in transportation,medical monitoring and other fields. Interest region is an active activity area. They can be railway stations,airports,schools,shopping malls and so on. A new algorithm for detecting trajectory outliers between interest regions,called DATIR(Detecting Anomalous Trajectories between Interest Regions),is proposed. Unlike the existing algorithm to detect from the local sample points,algorithm DATIR takes into account the local and global characteristics of the trajectory,and uses clustering method to detect the normal path and anomalous trajectory between interest regions. Algorithm DATIR has a wide range of application areas,such as taxi fraud monitoring,road planning,and so on. In order to improve the efficiency of mining trajectory from massive trajectory datasets,the parallel algorithm for detecting trajectory outliers based on MapReduce framework,which is called PDATIR(Parallel algorithm for Detecting Anomalous Trajectories between Interest Regions),is presented. The experimental results demonstrate that algorithm DATIR can effectively detect the anomalous trajectories between regions of interest,and can mine the normal path. Algorithm PDATIR shows obvious performance advantages over large data sets,and it has the high scalability and good speedup ratio.