Research On Dangerous Driving Behavior Recognition And Vehicle Tracking Algorithm Based On Data Fusion

With the rapid development of China's economy and the improvement of people's living standards,freight logistics industry has developed more rapidly.At the same time,road traffic safety issues have become increasingly prominent,one of the major problems for the industry department is how to reduce traffic accidents.Road traffic accidents statistics show that driver driving speeding,fatigue driving,driving distracted and other dangerous driving behavior is the main factor which causing road traffic accidents.According to statistics,the accident driving rate and mortality rate of the truck driving are higher than the bus,so the paper makes effective remote monitoring and real-time warning of dangerous driving behavior and vehicle driving state by means of technical means for the serious consequences of the dangerous driving behavior of the freight vehicle.The main content of this paper is to research the algorithms of dangerous driving behavior identification and vehicle tracking.At present,the commonly used fatigue detection method has the human eye fatigue recognition algorithm based on the eye characteristic parameters and the lane departure fatigue algorithm based on the lane departure characteristic parameters.However,the two kinds of fatigue detection algorithms have the problem of low accuracy.Therefore,according to the idea of data fusion,the rough set model is used to fuse the fatigue parameters based on human fatigue and the fatigue parameters based on the fatigue of the lane deviation.In the process of early warning of fatigue driving behavior,the vehicle tracking is to locate the vehicle,and then the long-distance transmission for the judgment result of fatigue driving behavior can be realised,to achieve the real-time warning of the vehicle driver's fatigue state.In order to better realize the long-distance monitoring and real-time warning,according to the Beidou/GPS double-mode positioning method,the Beidou system and the GPS system are solved by the information fusion algorithm based on the lossless Kalman filter.In this paper,the data fusion model based on the characteristic parameters of eye fatigue and the characteristics of lane deviation is established,which can effectively eliminate the data deviation and improve the accuracy of fatigue identification.Through vehicle tracking real vehicle remote monitoring and real-time warning,and the final location is obtained,and apply the dangerous driving behavior recognition and vehicle tracking research to the dangerous driving early warning process.