This paper proposed a real-time system to recognize and track multiple vehicles. The occlusion problem is solved by the occlusion segmentation method and propsoed recognition method uses the visual lenght, visual width, roof, windshield, and hood of vehicles to classify the vehicles to vans, utility vehicles, sedans, mini trucks, or large vehicle. Experiments obtained by using complex road scences are reported, which demonstrate the validity of the method in terms of robustness, accuracy, and time responses.
Real Time Recogniton and Tracking System of Multiple Vehicles
2006 IEEE Intelligent Vehicles Symposium ; 478-483
2006-01-01
4397080 byte
Conference paper
Electronic Resource
English
SMALL SPACECRAFT RECOGNITON USING RECURRENT NEURAL NETWORKS
TIBKAT | 2022
|LANE RECOGNITON DEVICE AND METHOD BASED ON LIDAR FOR VEHICLE
European Patent Office | 2022
|Real Time recognotion and Tracking System of Multiple Vehicles
British Library Conference Proceedings | 2006
|European Patent Office | 2019