智能车测试环境中虚拟车道线的快速构建方法

来源期刊:中南大学学报(自然科学版)2013年第3期

论文作者:卢远志

文章页码:985 - 991

关键词:机器视觉;视觉导航;车道线识别;智能车

Key words:machine vision; vision based navigation; lane detection; intelligent vehicle

摘    要:针对测试环境的设计与构建是智能车辆智能行为验证的前提,但常常会受到真实道路环境复杂性和不确定性等因素的限制等问题,提出序列化排列锥形交通标搭建智能车测试环境,在机器视觉基础上根据识别出的锥形标间的拓补关系构建虚拟车道线的方法。研究结果表明:该方法可有效地处理感知环境中的行人与运动车辆等干扰信息,满足智能车对系统的实时性和鲁棒性要求;该方法在结构化道路与非结构化道路均可实施,是一种快速构建智能车测试环境的有效方法,既降低了测试成本,又保证了测试过程的安全性和可靠性。

Abstract: Based on the fact that the testing environment is one of the significant and fundamental issues for vision based intelligent vehicle, but it is usually limited by the complication and uncertainty of the environment on practical road, in order to deal with the above issues, a rapid method on reconstruction of virtual lane environment for intelligent vehicle testing was presented. The traffic cones were ordered and arranged one by one on the road to construct the testing environment according to the requirements of performances verification, then the machine vision was used to detect and localize to generate the virtual lane according to their topological information. The results show that the identification of the traffic cones can be well done in the testing environment disturbed by the pedestrians and moving vehicles. The real-time and robust of the developed system can satisfy the requirements of the intelligent vehicles. The method can be applied in both structured and unstructured road, the testing safety can be guaranteed and the testing cost can be saved.

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号