Limits...
Mixed pattern matching-based traffic abnormal behavior recognition.

Wu J, Cui Z, Sheng VS, Shi Y, Zhao P - ScientificWorldJournal (2013)

Bottom Line: It introduced the DTW distance to measure the distances between vehicle trajectories and determined the number of clusters automatically by a spectral clustering algorithm based on the distance matrix.Then, it clusters sample data points into different clusters.The real-world application verified its feasibility and the validity.

View Article: PubMed Central - PubMed

Affiliation: The Institute of Intelligent Information Processing and Application, Soochow University, Suzhou 215006, China.

ABSTRACT
A motion trajectory is an intuitive representation form in time-space domain for a micromotion behavior of moving target. Trajectory analysis is an important approach to recognize abnormal behaviors of moving targets. Against the complexity of vehicle trajectories, this paper first proposed a trajectory pattern learning method based on dynamic time warping (DTW) and spectral clustering. It introduced the DTW distance to measure the distances between vehicle trajectories and determined the number of clusters automatically by a spectral clustering algorithm based on the distance matrix. Then, it clusters sample data points into different clusters. After the spatial patterns and direction patterns learned from the clusters, a recognition method for detecting vehicle abnormal behaviors based on mixed pattern matching was proposed. The experimental results show that the proposed technical scheme can recognize main types of traffic abnormal behaviors effectively and has good robustness. The real-world application verified its feasibility and the validity.

Show MeSH

Related in: MedlinePlus

Changing lane behavior recognition: (a) an abnormal trajectory, (b) the matching results of the spatial patterns, and (c) the matching results of mixed patterns.
© Copyright Policy - open-access
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3926328&req=5

fig7: Changing lane behavior recognition: (a) an abnormal trajectory, (b) the matching results of the spatial patterns, and (c) the matching results of mixed patterns.

Mentions: On highway, if a vehicle wants to change a lane, it should turn on signal in advance to call attention from the following car. But there always exists the abnormal lane-changing phenomenon, which easily results in traffic incidents. It is one of high-risking driving behaviors. In Figure 7(a), the motion trajectory marked with pink is a changing lane behavior. Figure 7(a) also shows the clustering results. Figure 7(b) shows spatial-pattern matching results. For the testing trajectory, at the beginning it traveled in the middle motion pattern, that is, the motion pattern marked in blue color. Therefore, the probability belonging to the blue pattern is 1, and the probability belonging to the red pattern and that of the green pattern are 0. With its gradual shift, the vehicle shifted to the green motion pattern and finally ran on the green motion pattern. Figure 7(b) shows that the probability of the blue motion pattern decreases gradually, downward to 0. The probability of the green motion pattern increases gradually, upward to 1. The probability of the red motion pattern always keeps as 0. Note that the sum of the probabilities belonging to the three patterns is 1. After integrating direction pattern matching with the basis of spatial pattern matching, the final abnormal recognition result is shown in Figure 7(c).


Mixed pattern matching-based traffic abnormal behavior recognition.

Wu J, Cui Z, Sheng VS, Shi Y, Zhao P - ScientificWorldJournal (2013)

Changing lane behavior recognition: (a) an abnormal trajectory, (b) the matching results of the spatial patterns, and (c) the matching results of mixed patterns.
© Copyright Policy - open-access
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3926328&req=5

fig7: Changing lane behavior recognition: (a) an abnormal trajectory, (b) the matching results of the spatial patterns, and (c) the matching results of mixed patterns.
Mentions: On highway, if a vehicle wants to change a lane, it should turn on signal in advance to call attention from the following car. But there always exists the abnormal lane-changing phenomenon, which easily results in traffic incidents. It is one of high-risking driving behaviors. In Figure 7(a), the motion trajectory marked with pink is a changing lane behavior. Figure 7(a) also shows the clustering results. Figure 7(b) shows spatial-pattern matching results. For the testing trajectory, at the beginning it traveled in the middle motion pattern, that is, the motion pattern marked in blue color. Therefore, the probability belonging to the blue pattern is 1, and the probability belonging to the red pattern and that of the green pattern are 0. With its gradual shift, the vehicle shifted to the green motion pattern and finally ran on the green motion pattern. Figure 7(b) shows that the probability of the blue motion pattern decreases gradually, downward to 0. The probability of the green motion pattern increases gradually, upward to 1. The probability of the red motion pattern always keeps as 0. Note that the sum of the probabilities belonging to the three patterns is 1. After integrating direction pattern matching with the basis of spatial pattern matching, the final abnormal recognition result is shown in Figure 7(c).

Bottom Line: It introduced the DTW distance to measure the distances between vehicle trajectories and determined the number of clusters automatically by a spectral clustering algorithm based on the distance matrix.Then, it clusters sample data points into different clusters.The real-world application verified its feasibility and the validity.

View Article: PubMed Central - PubMed

Affiliation: The Institute of Intelligent Information Processing and Application, Soochow University, Suzhou 215006, China.

ABSTRACT
A motion trajectory is an intuitive representation form in time-space domain for a micromotion behavior of moving target. Trajectory analysis is an important approach to recognize abnormal behaviors of moving targets. Against the complexity of vehicle trajectories, this paper first proposed a trajectory pattern learning method based on dynamic time warping (DTW) and spectral clustering. It introduced the DTW distance to measure the distances between vehicle trajectories and determined the number of clusters automatically by a spectral clustering algorithm based on the distance matrix. Then, it clusters sample data points into different clusters. After the spatial patterns and direction patterns learned from the clusters, a recognition method for detecting vehicle abnormal behaviors based on mixed pattern matching was proposed. The experimental results show that the proposed technical scheme can recognize main types of traffic abnormal behaviors effectively and has good robustness. The real-world application verified its feasibility and the validity.

Show MeSH
Related in: MedlinePlus