基于路径搜索的车道线检测算法

Lane detection algorithm based on path search

针对基于特征的车道线检测算法鲁棒性较差的问题,本文提出一种基于路径搜索的不依赖感兴趣区域的车道线检测算法.首先,对预处理图像平滑处理过滤部分噪声,并采用基于梯度的边缘分割方法提取车道线边缘;其次,提出一种基于路径搜索的车道线检测算法,生成两个包含车道线特征点的集合;最后,用三次多项式结合最小二乘法对所得车道线元素点进行曲线拟合,得到两侧车道线.试验结果表明,相比于基于Hough变换的车道线检测算法,该算法在校园道路场景下的鲁棒性更强,且在弯道车道线检测方面表现良好.

Aiming at the poor robustness of feature-based lane detection algorithm,a lane detection algorithm based on path search independent of region of interest is proposed in this paper.Firstly,the preprocessed images are smoothed to filter some noise,and the edge of lane line is extracted by using the edge segmentation method based on gradient.Secondly,a lane detection algorithm based on path search is proposed to generate two sets containing lane line feature points.Finally,the cubic polynomial combined with the least square method is used to fit the obtained lane line element points,and the lane lines on both sides are obtained.The experimental results show that compared with the lane line detection algorithm based on Hough transform,the proposed algorithm is more robust in campus road scene,and performs well in curve lane line detection.