基于Hough变换的车道检测改进算法研究  

Research of robust lane detection algorithm based on Hough transform

作　　者：李明[1] 黄华[1] 夏建刚[1] 

机构地区：[1]四川大学电气信息学院,四川成都610065

出　　处：《计算机工程与设计》2012年第4期1638-1642,共5页Computer Engineering and Design

摘　　要：车道检测是车道侧偏检测的前提,是车辆辅助驾驶系统的重要组成部分,针对这一特点,提出了利用Hough变换来实现对车道检测的直道检测,而对于弯道部分的检测,则是该算法的一个创新之处。通过统计较小区域的目标像素数目,设置感兴趣区域(ROI)来判断车道的弯曲方向,并在该方向上进行弯道像素的搜索,采用多段折线的方法对弯道部分进行重建。实验结果表明该算法具有运算速度快、准确性高、鲁棒性较好的特点。Aiming at lane detection, which is the premise of lane departure, plays an important roll in intelligent vehicle naviga- tion. Hough transform is proposed to achieve the straight lane detection. While, for the corner part of the lane detection, that is one of the paper＇s innovative points. Firstly, by statistic the number of pixels of small area, the interested region （ROD is set to determine the bending direction of the lane, and then the pixels of curve lane in the direction of corners pixels is searched, finally the corners parts of the lane rebuild by using multiline methods. Experimental results show that this method has a fast calcula tion speed, accuracy, high robustness.

关 键 词：车道检测 HOUGH变换 感兴趣区域 多段折线 偏离点 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]