摘要
车道线检测是自动驾驶系统和高级驾驶辅助系统的重要组成部分,为车辆提供自身的位置信息。为了提高检测系统的准确性和鲁棒性,本文提出了一种基于OpenCV实现的车道线检测方法,首先,针对摄像机采集图像产生的失真和畸变问题,利用黑白格子标定板对原始图像进行了标定,获得无畸变图像,然后对无畸变的RGB图像分别进行灰度化、平滑滤波、canny边缘检测、感兴趣区域的获取和霍夫变换等过程检测车道线。实验结果表明,该方法能够有效地解决图像中的光线明暗问题,车道线检测地准确率高达92. 49%,具有较高地准确性和鲁棒性,满足自动驾驶系统中对检测地实时性要求,在自动紧急制动系统(AEB)上具有较高地实用价值。
The lane line detection is an important part of the autopilot system and advanced driver assistance systems,providing the vehicle with its own location information. In order to improve the accuracy and robustness of the detection system,this paper proposes a lane line detection method based on OpenCV. First,the original image was calibrated to obtain an undistorted image,using a black and white grid calibration plate for the distortion problems generated by the camera acquisition images. And then these undistorted RGB images were processed with graying,smooth filtering,canny edge detecting,extracting the region of interest,and Hough transformation to detect lane lines. The experimental results show that this method can effectively solve the problem of light in the images. The detection accuracy of the lane line is as high as 92. 49%. It has high accuracy and robustness,and meets the real-time requirements of the detection in automatic driving systems with a high practical value.
作者
王文豪
高利
WANG Wenhao;GAO Li(Beijing Institute of Technology,Beijing 100081,China)
出处
《激光杂志》
北大核心
2019年第1期44-47,共4页
Laser Journal
基金
国家重点研发计划(No.2017YFC0804808)
关键词
车道线检测
标定
特征提取
lane line detection
calibration
feature extraction
