单目智能车道偏离预警系统  

Monocular Aptitude Lane Departure Warning System Design

作　　者：吕柯岩[1,2] 朱明[1] 贺柏根[1,2] 杨粤涛[1,2] 

机构地区：[1]中国科学院航空光学成像与测量重点实验室,中国科学院长春光学精密机械与物理研究所,吉林长春130033 [2]中国科学院研究生院,北京100049

出　　处：《液晶与显示》2012年第4期545-551,共7页Chinese Journal of Liquid Crystals and Displays

基　　金：中国科学院二期创新基本资助项目(No.C05T022)

摘　　要：车道偏离预警系统是继安全气囊之后的汽车安全辅助系统,该系统主要任务是采用基于机器视觉的方法提取车道线并进行预警决策。文章利用TMS320DM642视频处理器作为中央处理器,设计出基于DM642的车道偏离预警系统硬件架构,算法方面对图像进行灰度化、二值化和边缘提取做预处理,然后设置感兴趣区域(ROI),利用基于相位编组的改进Hough变换(RHT)进行车道线检测,根据车道偏离预警条件进行预警决策,当车辆在驾驶员非意识时偏离车道线的情况下实施报警。试验结果证明,本系统能够提前2.5s进行车道偏离的预警工作,并能够排除路面标记的影响,满足车道偏离预警系统实时、鲁棒的性能要求。Lane Departure Warning System（LDWS） is a car safety assist systems,following the capsule of the safety,whose main task is to extract lane markings and make warning-decision.In this paper,the lane departure warning system hardware architecture is designed based on TMS320DM642 which acts as the central processing unit.The algorithms turn the image into grayscale,and then detect the image＇s edge information,in the end set the region of interest（ROI） on the image.Hough transform based on phase grouping is used to extract the lanes,and according to the lane departure warning conditions,the system analyzes whether the warning should be presented or not and gives the driver warning singles when it is unconscious.The experiments results proved that the system will make warning in 2.5 s before the traffic accident,which implements the aim of the real-time,reliable lane departure warning.

关 键 词：车道偏离预警系统 TMS320DM642 HOUGH变换 相位编组 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]