01337765

Component

Visual-based traffic behavior analysis has attracted much attention in the field of intelligent transportation. Based on object trajectories data extracted by video tracking, motion patterns could be learned automatically, which is an effective approach for modeling and analyzing traffic behavior. Most current methods for motion pattern learning simply depend on one-sided characteristics of object trajectories which are not adaptable to the multiplicity and complexity of traffic activities. In this paper, we present a multi-level motion pattern learning approach for traffic behavior analysis, which is under taking into account spatial characteristic, direction characteristic and type characteristic of trajectories. At spatial-level, improved Hausdorff distance measurement is applied to construct spatial similarity matrix of trajectories collected and spectral clustering is used to realize spatial pattern learning. At direction-level, start points and end points of trajectories are fitted by 2-D GMM model to extract the distribution of entry and exit zones. Then, the direction pattern is obtained from the regional centers of the pairwise distribution zones. At type-level, the type pattern is acquired by K-means clustering algorithm with considering multiple classification features of trajectories. Based on the learned multi-level motion patterns, abnormal behavior detection algorithms are further developed by pattern matching. Finally, our approach is tested with several video sequences from real-world traffic scenarios. Experimental results show good performances of learning motion patterns. Some typical traffic behaviors in test scenes are successfully recognized and analyzed as well as examples of abnormal traffic behavior, lane changing and reverse driving are also reliably detected.

01329018

11-2222

English

16p

2011

Transportation Research Board 90th Annual Meeting

DVD

Figures (8) ; References (28)

Algorithms; Automatic vehicle detection and identification systems; Cluster analysis; Pattern recognition systems; Spatial analysis; Traffic surveillance; Travel behavior; Travel demand; Video imaging detectors

Data and Information Technology; Highways; I71: Traffic Theory

Transportation Research Board Annual Meeting 2011 Paper #11-2222

TRIS, TRB

Feb 17 2011 6:06PM