期刊文献+

车载单目摄像机下地铁轨行区检测与提取

Detection and Extraction of Rail Transit Area under Vehicle Mounted Monocular Camera
下载PDF
导出
摘要 地铁对线路运行区域(即轨行区)有非常严格的要求,异物侵入轨行区检测可有效提醒司机避免行车事故发生。轨行区域检测与提取作为轨行区异物自动检测中第一步具有重要的研究意义,在分析现有研究方法的基础上提出一种改进的轨行区检测算法。首先,对图像进行预处理完后利用canny算子进行边缘检测;然后,利用霍夫直线检测算法构建钢轨直线候选集;再次,利用直线斜率、直线端点偏移量等约束对钢轨直线候选集进行处理,移除异常数据并有效合并直线段;最后,用钢轨直线端点提取轨行区,为后续轨行区异物自动检测提供坚实基础。对某市地铁运行线路真实数据进行采集及实验,结果表明,本文所提方法可准确检测出地面区间、地面站台、隧道及地下车站等场景的轨行区,该方法提高了钢轨检测准确性和轨行区提取准确性。 Rail transit has very strict requirements on the line operation area(i.e.rail transit area).The detection of foreign matter intrusion into the rail transit area can effectively remind drivers to avoid traffic accidents.As the first step in the automatic detection of foreign objects in the track area,track area detection and extraction bears very important research significance.Based on the analysis of existing research methods,an improved track area detection algorithm is proposed in this paper.Firstly,Canny operator is used for edge detection after image preprocessing,and then Hough line detection algorithm is applied to construct rail line candidate set.In addition,constraints such as line slope and line endpoint offset are employed to process rail line candidate set to remove abnormal data and effectively merge line segments.Finally,the rail straight line endpoint is needed to extract rail line area,and it provides a solid foundation for the automatic detection of foreign matters in the track area.The real data collection and experiment of a metro line show that the proposed method can accurately detect the track area of the ground section,ground platform,tunnel and underground station,and improve the accuracy of rail detection and track area extraction.
作者 谭飞刚 余志立 刘开元 李汀 TAN Feigang;YU Zhili;LIU Kaiyuan;LI Ting(Shenzhen Institute of Information Technology,Shenzhen 518172,China)
出处 《铁道标准设计》 北大核心 2022年第9期74-77,共4页 Railway Standard Design
基金 广东省科技创新战略专项资金项目(pdjh2021b0907) 广东省普通高校青年创新人才项目(2020KQNCX205)。
关键词 车载摄像机 地铁轨行区 轨行区检测 轨行区提取 钢轨检测 vehicle mounted camera metro track area track area detection track area extraction rail inspection
  • 相关文献

参考文献14

二级参考文献112

  • 1吴艳华,何峰林,王富章,李平.基于灾害风险评估模型的铁路灾害监测预警系统研究[J].中国铁道科学,2012,33(1):121-125. 被引量:14
  • 2王志斌.哈大铁路客运专线雪深监测系统研究[J].铁道标准设计,2012,32(5):165-168. 被引量:6
  • 3高卫东,刘明哲,魏文寿,李景贤.铁路沿线风吹雪灾害及其防治研究[J].中国铁道科学,2004,25(5):97-101. 被引量:26
  • 4刘靖明,韩丽川,侯立文.基于粒子群的K均值聚类算法[J].系统工程理论与实践,2005,25(6):54-58. 被引量:122
  • 5Clements Mair, Dr Saeed Fararooy. Practice and potential of computer vision for railways[ C]. Condition Monitoringtbr Rail Transport. lEE Seminar, 1998 Nov. 10.
  • 6Jun Xue,Jun Cheng,Li Wang,et al. Visual monitoring- based railway grade crossing surveillance system [C]. 2008 Congress on Image and Signal Processing, 2008.
  • 7Susumu Kubnta, Tsuyoshi Nakano, Yasukazu Okamoto. A global optimization algorithm for Real-time On-board stereo obstacle detection systems [C].Proceedings of the 2007 IEEE Intelligent Vehicles Symposium lstanbul, Turkey, 2007, June 13-15.
  • 8-dayuki 'l-ugawa. Vision-based vehicles in Japan : Machine vision systems and driving control systems [J].IEEE Transactions on Industrial Electronics, 1994 vol. 41, August, 4.
  • 9Fatih Kaleli, Yusuf Sinan Akgul. Vision-based railroad track extraction using dynamic programming [ C ]. Proceedings of the 12th International IEEE Conference on intelligent Transportation Systems, 2009, Oct. 3-7.
  • 10Milan Ruder, Nikolaus Mohler, Faruque Ahmed. Anobstacle detection system [or automated train5[C]. Intelligent Vehicles Symposium, 2003.

共引文献138

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部