基于激光雷达+视频的城际铁路周界入侵监测报警技术研究

Detection and Alarm Technology of Intercity Railway Perimeter Intrusion Basedon Laser Radar and Video

城际铁路依靠标准化的作业流程和封闭式的物理护墙来规范旅客行为,然而存在少部分旅客异常行为和异物侵限事件对列车的正常运营造成安全风险,甚至引发安全事故。通过调研城际铁路典型运行环境特点和各类周界入侵技防手段特点,提出激光雷达+视频融合的城际铁路周界入侵监测技术方案。在考虑场景和技术路线适配性的同时,分析气象条件和外部环境对监测技术的影响,通过对烟尘雨雾小颗粒过滤算法、ICP雷视坐标配准算法、点云目标识别算法的针对性研究,提高在雨雾天气的适用性和识别准确率;此外,对雷视融合算法进行研究,提出基于像素级的数据融合技术和基于决策级的报警技术,提高检测结果的准确率,同时提升检测效率,增强目标检测算法的鲁棒性。

Intercity railway relies on standardized operation process and closed physical wall to regulate passenger behavior,whereas a small number of passengers’abnormal behavior and foreign body intrusion incidents cause security risks to the normal operation of trains,and even lead to safety accidents.Based on the investigation of typical operating environment characteristics of intercity railways and the characteristics of various perimeter intrusion technical prevention methods,this paper proposed a laser radar and video fusion technology for intercity railway perimeter intrusion monitoring.The influence of meteorological conditions and the external environment on the monitoring technology was analyzed while considering the adaptation of scenarios and technical routes.The applicability and recognition accuracy in rain and fog weather were improved through the targeted research on small particle filtering algorithm,ICP camera–radar coordinate registration algorithm and point cloud object recognition algorithm.In addition,research on the camera–radar fusion algorithm was proposed,and a pixel-level data fusion technology and a decisionlevel alarm technology were proposed to improve the accuracy of detection results,improve detection efficiency,and enhance the robustness of the object detection algorithm.