基于视频的智能交通异常检测系统

A traffic abnormity detection algorithm based on vision in the ITS

交通异常情况检测一直是交通管理中的重要任务,其在智能交通系统中显得尤为重要。传统的检测方法存在许多不足,不能满足视频监控的需要。基于此,提出一种基于视频图像处理的异常情况检测算法,在对视频图像进行非线性校正后,通过基本的图像处理获得较准确的交通信息,对各时刻的交通状态进行判断并可对下一阶段的状态进行预测。实验结果表明,该算法具有较强的有效性,在ITS中有较好的应用。

The traffic abnormity detection is an important task in the traffic management system, and it becomes more important in the ITS. The traditional detection algorithms have several deficiencies and can not satisfy the visual surveillance system. An algorithm based on digital image processing can get the accurate parameters of traffic by rectifying the nonlinear distortion. This algorithm can detect and predict the traffic abnormity. The results of experiments show that this algorithm has availability and has good applications in the ITS.