We propose a macroscopic method, which can perform real-time tracking of moving vehicles on highway. In addition to dealing with the problem caused by occlusion and noise, the proposed method can also track a vehicle doing lane change. The proposed method consists of two phases: the detection phase and the tracking phase. In the detection phase, we use entropy-based features to check the existence of vehicles. Then, we use the flux theory, which is commonly used in fluid mechanics to perform the tracking task. By conducting a great number of experiments, we have demonstrated the efficiency as well as the effectiveness of the proposed system.
Real-time vehicle tracking on highway
2003-01-01
424965 byte
Conference paper
Electronic Resource
English
Real-Time Vehicle Tracking on Highway
British Library Conference Proceedings | 2003
|Real-time tracking highway traffic safety early warning system
European Patent Office | 2023
|