一种基于视频图像的道路检测方法

Approach for road detection based on video image

实现车辆自动导航必须先进行道路检测,即获取道路相对于车辆的位置、形状等几何信息。本文算法首先预测图像中道路所在区域,使得需处理的数据量大大减少;然后使用Canny算子检测道路边缘,但对Canny算法进行了一些修改,以提高算法速度;最后使用Hough变换估计出道路直线的参数。对于初始化问题,则通过结合各种先验知识来解决。使用在北京街道上随车拍摄的交通视频做实验,结果表明该算法能在大多数情况下实时、准确地跟踪检测车道。

It's necessary for automatic vehicle navigation to detect the road first,i.e.to get the geometric information of the road.In our approach,the location of the road is estimated first,which can reduce the amount of data dramatically.Then the edge of the road is detected by using modified Canny detector in order to improve the efficiency.At last the parameters of the road model are identified by the Hough transform.Initialization is implemented with the help of prior knowledge.Video images obtained by an in-vehicle camera are utilized to the experimental road detection.The results show the feasibility of the approach with applications to real-time and precise detection of most structured roads.