摘要
针对结构化道路车道线检测的鲁棒性与实时性问题,本文提出一种基于概率霍夫变换的车道线识别方法。为了提高检测效率,对感兴趣区域内的图像利用概率霍夫变换方法检测出直线段;用最小二乘法拟合出正确的车道线并计算消失点的坐标,通过计算消失点横坐标和图像中心点横坐标的位置关系预测车道线方向;利用Visual Studio 2013平台完成车道线识别仿真验证。仿真结果表明,本文算法可以在不同环境下检测出车道线特征,有效提高了结构化道路车道线检测的鲁棒性与实时性。
In order to improve the robustness and real-time performance of structured road lane detection,a lane line recognition method based on probabilistic Hough transform is proposed.After thresholding the image,Canny edge operator is used to detect the edge of lane line.After that,the image in the region of interest is processed,the straight line is detected by using the probabilistic Hough transform method,and the correct lane line is fitted by the least square method.Then the coordinates of the vanishing point according to the fitted straight line are calculated,the lane line direction by calculating the position relationship between the abscissa of the vanishing point and the abscissa of the image center point is predicted,and finally the lane line recognition with the help of visual studio 2013 platform is completed.The verification results show that the algorithm can detect the lane features in different environments.
作者
龚鹏
张晏悦
GONG Peng;ZHANG Yanyue(Shenyang Aerospace University,Mechanical and Electrical Engineering,Shenyang Liaoning 110136,China)
出处
《汽车零部件》
2022年第S02期51-56,共6页
Automobile Parts
基金
辽宁省“百千万人才工程”人选科技活动支持项目(2020921030)。
