摘要
汽车追尾预警是智能车辆视觉导航系统中的重要研究内容。设计了一个应用于结构化道路环境,基于单目视觉的汽车追尾预警系统。该系统按照摄像机提供的道路图像序列,首先利用一种新的边缘检测算法识别前方道路,然后利用灰度、边缘和对称性等特征识别前方车辆,接下来根据前后车距判断其威胁等级,最终向驾驶员提供相应的声光报警信号。该系统已在合肥的高速公路上进行了实验。实验结果显示,系统运行速度达到车辆驾驶的实时性要求,能够完成车辆检测和碰撞预警的任务。
Rear-end collision warning is an important research topic of intelligent vehicle vision navigation system.A monocular camera based rear-end collision warning system in car was developed,which can be applied to the structuralization road environment.The front road is identified by a new edge detection algorithm from images captured by camera..The front cars are detected according to the characteristics such as shade of grey,edge and symmetry.The danger degree is judged according to the distance between cars.Acoustic and ocular warning signals are given to drivers.Experiment was conducted in highway around HeFei.The experimental results show that the system can meet the requirements of real-time for vehicle driving and can fulfill tasks of detection and warning.
出处
《现代电子技术》
2011年第20期53-56,共4页
Modern Electronics Technique
关键词
车辆检测
碰撞预警
单目视觉
智能车辆
car detection
collision warning
monocular vision
intelligent vehicle
