基于背景重构的公路车流量检测方法

在分析现有检测算法的基础上,提出了一种车流量统计算法,应用自适应背景重构方法,首先利用差分法提取动目标,然后通过先边缘检测再膨胀腐蚀的方法去除车影,最后在虚拟检测线上分区域段设置资源信号灯统计车流量。实验结果表明,本算法能够适应公路背景的变化,有效地去除了车影,统计结果准确可靠。

A Vehicle Flow Detection Method Based on Background Reconstruction

An algorithm of vehicle flow statistics was presented based on analyzing the existed algorithm,and the automatic adaptive background reconstruction method was utilized.It withdrawed the moving object by difference method,and adopted the vehicle shadow removing method by first fringe detection and second dilation and erosion.The vehicle flow was calculated by setting resource signal lamp in each segment of dummy detection line.The experiments confirm that,this algorithm can adapt the change of the background of road,and it can remove the vehicle shadow effectively,and the statistics result is exact and reliable.