[1]李 玉,王桂丽,张道秧,等.一种交通路口车道线检测与车道分割方法[J].计算机技术与发展,2022,32(S2):72-76.[doi:10. 3969 / j. issn. 1673-629X. 2022. S2. 013]
　LI Yu,WANG Gui-li,ZHANG Dao-yang,et al.A Lane Line Detection and Segmentation Method for Traffic Intersections[J].,2022,32(S2):72-76.[doi:10. 3969 / j. issn. 1673-629X. 2022. S2. 013]
点击复制

一种交通路口车道线检测与车道分割方法()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
32
期数:
2022年S2期
页码:
72-76
栏目:
图形与图像
出版日期:
2022-12-11

文章信息/Info

Title:
A Lane Line Detection and Segmentation Method for Traffic Intersections
文章编号:
1673-629X(2022)S2-0072-05
作者:
李 玉1 ; 王桂丽1; 2 ; 张道秧1 ; 汪慧兰1; 2 ; 张保俊1 ; 刘志祥1
1. 安徽师范大学 物理与电子信息学院,安徽 芜湖 241000;
2. 智能机器人信息融合与控制研究所,安徽 芜湖 241000
Author(s):
LI Yu1 ; WANG Gui-li1; 2 ; ZHANG Dao-yang1 ; WANG Hui-lan1; 2 ; ZHANG Bao-jun1 ; LIU Zhi-xiang1
1. School of Physics and Electronic Information,Anhui Normal University,Wuhu 241000,China;
2. Institute of Intelligent Robot Information Fusion and Control,Wuhu 241000,China
关键词:
边缘检测算子; 霍夫变换算法; 车道线检测; 车道线合并; 车道分割
Keywords:
edge detection operator; Hough transform algorithm; lane line detection; lane merging; lane segmentation
分类号:
TP301;TN919. 8
DOI:
10. 3969 / j. issn. 1673-629X. 2022. S2. 013
摘要:
对交通路口所有车道上车辆检测并计数是实时调控交通信号灯配时的依据,而对不同车道上车辆准确计数的前提是实现车道准确分割。 因此,该文主要研究多车道线的检测与分割技术。 首先,针对交通路口车道线多、车辆密集等复杂场景,设计霍夫变换投票阈值,以去除大量干扰线;其次,根据斜率相似原理,采用端到端斜率比较算法,把多个共线的线段连成一条直线;最后采用延长算法,把车道线延长至图像边界,以实现对交通路口的车道线完整分割。 实验结果表明,该算法对于具有车辆密集、车辆重叠度高的交通路口的图片,可以准确进行车道线检测并且分割车道。
Abstract:
The real-time timing of traffic lights is based on the number of vehicles on each lane of the intersection. The premise ofaccurate counting of vehicles on different lanes is to achieve accurate lane segmentation. Therefore,we mainly study the detection andsegmentation technology of multilane lines. First of all,we design the voting threshold of Hough transform to remove a large number ofinterference lines,which are generated by complex scenes such as many lane lines and dense vehicles. Secondly, according to theprinciple of slope similarity,the end-to-end slope comparison algorithm is used to connect multiple collinear lines into a straight line. Finally,the lane line is completely segmented by extending the lane line to the image boundary. The experimental results show that the algorithm can accurately detect lane lines and segment lanes for the images of traffic intersections with dense vehicles and high overlap.
更新日期/Last Update: 2022-10-10