Video-Based Traffic Parameter Extraction with an Improved Vehicle Tracking Algorithm
Songhao Wang, Zhiheng Li, Zuo Zhang, Ji-Yuan Tan · 2012
Traffic parameters are the basis of traffic flow research. Video-based traffic parameter extraction, which can provide large amounts of information at a relatively lower cost, has attracted great interest in recent years. In this paper, we propose an improved vehicle tracking algorithm to automatically extract traffic parameters over a long-distance urban freeway section from a monocular traffic image sequence. The algorithm can overcome the limitation of occlusions and omissions. It includes background estimation, target segment, vehicle tracking and parameter collection. Test results show the robustness and effectiveness of the algorithm: over 95% of the vehicles can be extracted from the video accurately.