一种基于路侧摄像头的车辆跟踪方法

A Method of Vehicle Tracking Using Roadside Camera

提出了一种基于路侧摄像头的车辆跟踪方法。该方法通过在HSV空间实现阴影抑制降低阴影现象对ViBe算法的影响,实现车辆检测。在车辆跟踪阶段,通过卡尔曼滤波算法与改进ViBe算法结合,建立弱重叠视域的车辆跟踪模块,实现了跨监控区域的车辆跟踪。实验结果表明,该方法车辆检测及跟踪效果好,且能提取目标车辆运动信息,将运动目标信息发送出去,满足实际道路使用需求。

In this paper, a method of vehicle tracking using roadside camera is proposed. This vehicle detection method can reduce the interference of shadow on the ViBe algorithm by shadow suppression in the HSV color space. A vehicle tracking module with weak overlapped field of view is established in the vehicle tracking stage through the combination of the Kalman filtering algorithm and the improved ViBe algorithm to realize vehicle tracking across monitoring areas. The result of an experimental study shows that the method has good detection and tracking effects, and can extract the motion information of the target vehicle and transmit it to meet the actual road use requirements.