Clustering Vehicle Trajectories with Hidden Markov Models
Application to Automated Traffic Safety Analysis

International Joint Conference on Neural Networks, IEEE. Vancouver, 2006

Nicolas Saunier and Tarek Sayed

Intelligent Transportation Systems need to automatically monitor the road traffic, and especially track vehicles. Most research has concentrated on highways. Traffic in intersections is more variable, with multiple entrance and exit regions. This paper describes a extension to intersections of the feature-tracking algorithm described in \cite{beymer97real-time}. Vehicle features are rarely tracked from their entrance to their exit of the field of view. Our algorithm can accomodate this problem. It is evaluated on video sequences recorded on four different intersections.

Keywords: intelligent transportation systems, vehicle tracking, features, intersection

[bib] [pdf] [slides pdf]

Click on the image to see a demonstration video of traffic conflict detection in a sequence.