一种改进的车道线识别算法

An Improved Lane Recognition Algorithm

车道线识别是车辆辅助驾驶系统中重要组成部分。针对传统Hough变换,文章做出了改进,首先,对道路图像进行感兴趣区域划分、灰度化、二值化、中值滤波、边缘检测预处理,然后考虑到车道线满足一定角度和长度的限制,对车道线进行极径极角的约束,最后实现车道线提取。实验表明该方法具有较好的实时性和鲁棒性。

Lane recognition is an important part of vehicle auxiliary driving system.In this paper,the traditional Hough transform is improved.Firstly,the road image is preprocessed by region of interest partition,grayscale,binarization,median filtering and edge detection.Then,considering that the lane line meets certain angle and length constraints,the lane line is constrained by radius polar angle.Finally,the lane line is extracted.Experiments show that the method has good real-time performance and robustness.