Trajectory‐based anomalous behaviour detection for intelligent traffic surveillance. Issue 8 (1st October 2015)

• Record Type: Journal Article
• Authors: Cai, Yingfeng; Wang, Hai; Chen, Xiaobo; Jiang, Haobin
• Is Part Of: IET intelligent transport systems. Volume 9:Issue 8(2015)
• Page Start: 810
• Page End: 816
• Publication Date: 2015-10-01
• Subjects: intelligent transportation systems -- video surveillance -- learning (artificial intelligence) -- road vehicles -- hidden Markov models -- traffic engineering computing; trajectory based anomalous behaviour detection -- intelligent traffic surveillance -- trajectory analysis -- trajectory pattern learning module -- online abnormal detection module -- coarse‐to‐fine clustering strategy -- vehicle trajectories -- main flow direction -- MFD vectors -- filtering algorithm -- robust K‐means clustering algorithm -- coarse cluster -- hidden Markov model -- HMM -- path pattern -- online detection module -- vehicle trajectory -- MFD distributions -- motion patterns -- abnormal behaviour -- intelligent surveillance applications; Intelligent transportation systems -- Periodicals; Electronics in transportation -- Periodicals; 388.31205
• DOI: 10.1049/iet-its.2014.0238
• Languages: English
• ISSNs: 1751-956X

Abstract

Abstract : This study proposes an efficient anomalous behaviour detection framework using trajectory analysis. Such framework includes the trajectory pattern learning module and the online abnormal detection module. In the pattern learning module, a coarse‐to‐fine clustering strategy is utilised. Vehicle trajectories are coarsely grouped into coherent clusters according to the main flow direction (MFD) vectors followed by a three‐stage filtering algorithm. Then a robust K ‐means clustering algorithm is used in each coarse cluster to get fine classification by which the outliers are distinguished. Finally, the hidden Markov model (HMM) is used to establish the path pattern within each cluster. In the online detection module, the new vehicle trajectory is compared against all the MFD distributions and the HMMs so that the coherence with common motion patterns can be evaluated. Besides that, a real‐time abnormal detection method is proposed. The abnormal behaviour can be detected when happening. Experimental results illustrate that the detection rate of the proposed algorithm is close to the state‐of‐the‐art abnormal event detection systems. In addition, the proposed system provides the lowest false detection rate among selected methods. It is suitable for intelligent surveillance applications.