Research On The Key Technologies Of Vehicle's Overspeed Warning System In Highway

With the development of the national economy, the traffic mileage of highway, sizes and scales of motor vehicles is experiencing a substantially increasement. Enhancing the level of highway service, assorting with and inducing traffic to protect the safety of Secretary and reduce the accident rate in road accidents, set up speeding alert system and supply the driver of the vehicle with warning signs in the accident-multiple road section so that drivers can slow down in the road ahead of time to avoid a matter of urgency caused by overspeeding. Intelligent real-time traffic monitoring is of significant to traffic information collection, standardization of traffic management and traffic accidents decrease and so on .In this paper,the composing and work flow of the highway traffic overspeed warning system are studied the theory and method of the video speed measuring is expatiated, the production, characters and effects of vehicle's shadow of video speed measuring are also discussed. The method of vehicle tracking, the characters selecting, and the errors of video speed measuring are analyzed. The effect of the shadow of vehicles to the speed measuring is studied, the problems of vehicle's shadow and vehicle tracking are discussed. The solution to shadow testing and eliminating of the vehicle video testing, the vehicle locating and tracking of the, the vehicles speed accurate measuring and other related issues are presented. In accordance with the basis of video image processing, image filtering method, the algorithms of detecting shadow and vehicle tracking are systematically studied, and the improvement of the existing system are summarized, and the algorithm which is fit to realtime, dynamic shadow elimination and t vehicles tracking are proposed. The multi-scale Retinex algorithm with the Kimmel variation are proposed, experimental analysis are carried out, and Kalman filter is employed for image filtering, tracking and the vehicle identification by the Cam-Shift tracking methods, as a result the vehicle is tracked accurately and realtimely.