摘要
针对路面车道线的磨损现象,提出了一种精确的车道线检测方法,以解决磨损车道线的检测定位问题。通过把Hough投票看作是一个统计变量,计算统计均值和统计方差,基于统计特征计算车道线的距离、方向。此外,直接在灰度或二值图像上进行处理,避免了边缘检测过程。实验表明,当车道线磨损严重时,该方法也能准确地检测出车道线。
In view of the lane wear phenomenon,this paper proposes a robust method for lane detection in order to solve the problem of detecting and positioning the worn lane.By taking the Hough vote as a statistical variable,the statistical means and variances are computed.The lane parameters such as distance and direction are obtained based on the statistical characteristics.The edge detection process is avoided by directly processing the gray or binary image.The experimental results show that this method can accurately detect the lane in the presence of worn pavement.
作者
辛晓梦
尹雪
赵宇航
曹旦华
徐则中
XIN Xiaomeng;YIN Xue;ZHAO Yuhang;CAO Danhua;XU Zezhong(School of Computer Information Engineering, Changzhou Institute of Technology, Changzhou 213032)
出处
《常州工学院学报》
2020年第6期37-40,共4页
Journal of Changzhou Institute of Technology
基金
国家自然科学基金项目(61602063)
江苏高校文化创意协同创新中心立项课题(XYN1705)。
关键词
车道线检测
霍夫变换
统计特征
lane detection
Hough transform
statistical characteristics
