汽车超载监测系统

为了实时识别各种车型的超载车辆,该系统基于开源计算机视觉库(OpenCV),先根据车辆照片库建立车型分类器,然后使用数字摄像机拍摄进入监控区域的车辆,在视频中使用分类器识别车型,根据所识别得到的车型去查询数据库获得该车型的核载,再通过动态称重技术获得车辆的实际载重,及时判别车辆是否超载。此方法可避免过去使用统一重量衡量不同车型是否超载的弊端,并可同时免线圈测量车速。测试结果表明系统能快速准确地识别出车型。配合动态称重系统,就能实时得出所通过的车辆是否超载,对公路养护和道路交通安全有相当大的实用意义。

Monitoring system for overload vehicles

This system based on the open source computer vision library（OpenCV）is used to execute the real-time iden-tification of overloading vehicles with different vehicle models. A vehicle model classifier is established according to the vehicle photo library,and then the vehicles entering the monitored area are recorded by a digital camera. The vehicle models are identi-fied in video recognition by the classifier. The rated load of a vehicle model is obtained by querying the database,and then the actual load of the vehicle running across the monitoring area is measured by the dynamic weighing technology for judging if the vehicle is overloaded immediately. This approach can avoid the demerit of the previous one that uses the unified weight to esti-mate whether vehicles with different models are overloaded or not. In addition,it can also avoid the coil measurement of speed. Testing results show that the system can quickly and accurately identify a vehicle model,and can promptly determine whether the vehicle is overloaded by means of dynamic weighing system. It has considerable practical value for road maintenance and road traffic safety.