Vehicle Detection Algorithm Based On Wireless Sensor Networks

Intelligent Traffic System (ITS) is the scientific and effective method tosolve the city’s traffic problem. It mainly includes that traffic informationcollection, transmission and processing of traffic information. It improves itsefficiency by increasing the effective use and management of the trafficinformation. Vehicle information detection is an important part of intelligenttraffic system, and also it is the foundation and the key technology of ITS.Wireless Sensor Networks (WSN) is a new technology which is theintegration of wireless communication technology, microelectronics sensor, andthe embedded system. It can provide an effective mean for the acquisition andtransmission of information, monitor the traffic flow in each direction, speed andalso the status of the parking plot to finish the process of traffic informationacquisition, transmission and processing. Vehicle detection system mainly basedon WSN includes vehicle monitoring system, highway vehicle detection systemand intelligent parking space detection system.This paper presents a vehicle detection algorithm which is based on theabove three systems, with the help of the geomagnetic sensor we can detect theinformation of vehicles,such as the length of the vehicle, vehicle velocity, roadtraffic, occupancy of road, parking utilization and parking status. Vehicledetection algorithm based on WSN includes road vehicle detection algorithm,highway speed detection algorithm and parking space state detection algorithm.Road vehicle detection algorithm which means subdividing processing to vehiclesignal, presents a method to use the arithmetic mean and median hybrid filter toeliminate noise signal and obtain smooth vehicle signal. According to thedifferential method, we can obtain the related information of the road vehicles;accurately calculate the current road traffic information, road occupancy rate.Internet connection can transmit the city’s road traffic situation to users in realtime, which is obviously very convenient to users. A large number of experimentsverify the feasibility and robustness of the algorithm in this paper.Highway speed detection algorithm includes single node speed estimationand double node speed detection. It presents a statistical method to obtain singlenode road vehicle flow velocity estimation and double node vehicle speedestimation. Especially in the two-node synchronization algorithm thecompensation principle had been added, which can improve the two-node synchronization problem, and calculate the speed and length of the vehicle.In the parking lot, the node use finite difference method to obtain theparking state information, the computer uses a decision algorithm to obtain theparking state information. In this way we can detect the parking plot stateefficiently and accurately. The connection of the induction system can achievelarge parking intelligent management scheme.