摘要
针对传统车道检测方法不能同时满足检测精度和实时性要求的问题,文章提出了一种改进的基于密度的聚类方法对车道边沿点进行检测,然后利用最小二乘法拟合出车道边沿线。与传统DBSCAN算法相比,车道边沿检测精度提高了22.8%,平均检测时间降低14.2%,改进的DBSCAN车道检测算法具有更好的检测精度和实时性。
In view of the fact that the traditional lane detection method cannot simultaneously meet the detection accuracy and real-time requirements,an improved density-based clustering method is proposed in the paper to detect the lane edge points,and then the least square method is used to fit the lane edge lines.Compared with the traditional DBSCAN algorithm,the improved DBSCAN algorithm increases the detection accuracy by 22.8%and reduces the average detection time by 14.2%,which has better detection accuracy and real-time performance.
作者
李雨生
胡广地
LI Yusheng;HU Guangdi(Institute of Automotive Engineering,Southwest Jiaotong University,Chengdu 610031,China)
出处
《电工技术》
2020年第10期20-22,共3页
Electric Engineering
