复杂道路条件下的车道线检测算法  

Lane detection algorithm under complex road conditions

作　　者：姚洪涛[1] 张海萍 郭智慧 YAO Hongtao;ZHANG Haiping;GUO Zhihui(College of Computer Science and Technology,Changchun University of Science and Technology,Changchun Jilin 130000,China)

机构地区：[1]长春理工大学计算机科学技术学院,长春130000

出　　处：《计算机应用》2020年第S02期166-172,共7页journal of Computer Applications

基　　金：吉林省发展和改革委员会项目(2015G010)。

摘　　要：针对在复杂道路条件下基于滑动窗口的车道线检测算法易发生误检、漏检的现象,提出了一种改进算法。该算法在图像预处理方面使用了多阈值过滤方法,主要通过结合梯度阈值过滤和颜色空间阈值过滤的优势,充分提取道路图像中车道线的信息,消除了由复杂路面引起的噪声干扰;在车道线拟合方面,使用了基于动态自适应感兴趣区域的车道线跟踪方法,该方法主要利用前一帧的车道线信息自动调整感兴趣区域的位置及大小,有效地解决了基于滑动窗口的车道线检测算法在复杂行车环境下容易出现漏检、误检的问题。最后通过与基于滑动窗口和Hough变换的车道线检测算法进行对比实验,结果显示:改进后的车道线检测算法在准确性与实时性方面具有明显优势。Aiming at the phenomenon that the lane detection algorithm based on sliding window is prone to false detection and missed detection under complex road conditions,an improved algorithm was proposed.In the aspect of image preprocessing,a multi-threshold filtering method was used,which mainly combined the advantages of gradient threshold and color space threshold to fully extract the lane information from road images,the noise interference caused by complex road surfaces was eliminated.In terms of lane fitting,a lane tracking method based on dynamically adaptive regions of interest was used.In this method,the lane information of the previous frame was mainly used to automatically adjust the position and size of the region of interest,which effectively solved the problem that the lane detection algorithm based on sliding window was prone to missed detection or false detection in complex driving environment.Finally,a comparative experiment with the lane detection algorithm based on sliding window and Hough transform shows that the improved lane detection algorithm has obvious advantages in accuracy and real-time.

关 键 词：车道线检测 多阈值过滤 动态自适应 感兴趣区域 车道线跟踪 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]