(ITS) is the direction of dozens of traffic directions. Its development and development of computer technology, sensor technology, information technology and communication technology, etc. are effectively integrated and applied to Yu Yu Yu Yu. Large, real-time and accurate integrated integrated management system. . In the information flow of traffic information, all-round real-time traffic data in the information flow can be integrated and analyzed. When driving, the vehicle can be out of date; at the same time, the analysis of data can be It is possible to leave the car without overloading.
1 Traffic Safety

For fatigue driving, one of the most noteworthy problems in traffic safety, the driver's pupil can be located and tested by combining an infrared camera and a pulsed light-emitting diode. After image processing and feature extraction, it can be analyzed whether the pupil becomes smaller or even closed. Analyze the driver's fatigue driving situation; at the same time, collect the change information of the vehicle's driving direction through the axle steering angle sensor, and analyze whether the vehicle deviates from the driving lane; through the millimeter-wave radar installed in front of the vehicle, the vehicle and the vehicle in front or obstacles can be obtained in real time Distance, when the distance exceeds a certain threshold range, an alarm display and alarm sound will be issued, and at the same time, emergency braking will be taken when the driver has no time to brake to prevent car accidents. The system structure is shown in Figure 1.
2 Traffic flow information detection and fusion

Traffic flow information detection is an important part of intelligent transportation. After the information detected by each sensor is fused, real-time traffic flow data, such as traffic flow, vehicle speed, lane occupancy, traffic flow density, queue length, etc., are sent to the control center. After analyzing traffic flow information, corresponding instructions are issued to realize intelligent control. These parameters are mainly detected by loop coil detectors installed on the road surface. The principle of loop coil detection of traffic flow information is based on the inductance change of the loop. When a vehicle passes the coil position, the detector will be triggered to output a signal. When the vehicle leaves the coil, the signal will weaken, so the vehicle passes the coil detector. Waveform signal, counting the waveform will get the traffic flow through the coil in a certain period of time. Two coils are installed every 3 to 5 meters on the road. When the vehicle touches coil A, the pulse counts. When the vehicle touches coil B, the pulse count ends. In this way, the pulse counting time required for the vehicle to pass this distance can be obtained. The speed of the vehicle can be measured.
In the stage of multi-information fusion, first assume that the vehicle speed obeys the normal distribution, and use the weighted average data fusion method to fuse the vehicle speed parameters; in the traffic flow detection part, the mathematical model of the traffic flow distribution is determined according to the vehicle speed information obtained by fusion, That is, the vehicle speed is less than 24.5m/s, the traffic is congested, and the traffic flow obeys the binomial distribution, otherwise the traffic flow obeys the Poisson distribution; in the queue length prediction part, the estimated traffic flow information and the queue length at the previous moment are jointly used as Carl The parameters of the Mann filter prediction model are used to obtain the queue length prediction at the current moment.
3 Vehicle Tracking and Positioning

The working principle of the car anti-theft alarm system is: when the owner uses RFID (radio frequency identification technology) to turn on the anti-theft monitoring system, the entire anti-theft system starts to work, the strain sensor installed on the door can monitor the opening and closing of the car door, and the glass breakage detector can detect Whether the glass is deliberately damaged, and the pyrosensor can monitor whether someone has entered the car, and the Hall sensor can monitor the vibration of the car body. These sensors work together, and then transmit the information to the central processing unit for analysis in real time, so as to make a decision on whether the car body is safe. If it is judged that the car body has been invaded, the alarm module will send an alarm message to the mobile phone through GPRS to notify the car owner. On the other hand, the execution module can trigger the lights to flash or whistle to attract the attention of people around.
4 Conclusion

Multi-sensor fusion is an emerging technical field, and its rapid development provides a good solution for information extraction and analysis in the transportation field. Multi-sensor fusion can reasonably coordinate multi-source data to the greatest extent, make full use of useful information, and improve the comprehensive utilization of information. This advantage can be fully utilized in the field of intelligent transportation. The research on traffic information fusion technology has just begun, and there are still many problems that need to be solved urgently. The recognition reliability of targets and high-speed moving targets, etc., for a long period of time in the future, there will be more intelligent transportation technologies and fusion algorithms.