Traffic congestion identification based on image processing

Accurate and real-time traffic information is the foundation of intelligent transportation systems (ITS). In general, density, velocity and flow are used to describe traffic status of certain road segment. However, these macroscopic parameters are not able to reflect detailed traffic scenarios. It is more valuable to detect traffic congestion, which can be the basis of dynamic control and real-time guidance. This study proposes a novel approach towards traffic congestion identification based on vehicle trajectories in intelligent vehicle infrastructure co-operation system (IVICS). Considering spatial-temporal trajectories as image, this study uses self-correlation to extract propagation speed of congestion wave. Based on this, this study constructs congestion template; by matching algorithm, congestion is further identified as well as its intensity. Simulations on next generation simulation (NGSim) dataset verify the effectiveness of the above methods.