Vehicle tracking for traffic analysis
Researchers from the Augmented Vision group at DFKI have developed a novel approach for automatic video-based traffic analysis. First, entry and exit zones of complex crossroads or roundabouts are defined interactively. Then, each car is tracked individually by computing the trajectory based solely on the video content. This provides unique data for traffic analysis and in particular for supporting simulations for improved traffic solutions.