多源数据互补的隧道运营安全智能监测系统设计与应用

Design and Application of Multi-Source Data Complementary Intelligent Monitoring System for Tunnel Operation Safety

隧道作为高速公路重要的一环,以“人-车-环境”为主要因素的运营安全问题一直为管理者所关注。通过系统平台的架构与功能设计,以基于深度学习的视频事件检测技术为基础,适当增加如“雷达+视频”的融合检测技术、隧道态势感知技术、智能球机联动功能,提高了异常事件的检出效率,可掌握隧道实时运行状态以及事后的智能聚焦查看。解决了依靠人工巡检视频监控隧道交通事件的难题,其智能化、及时性、准确性对于预防二次事故的发生和控制事态发展,具有重要作用。

As an important part of the highway,tunnel operation safety with“man-vehicle-environment”as the main factor has always been concentratedonby the managers.Through structural establishment and function design of the system platform,based on the video event detection technology on deep learning,in this paper,such technologies have beenappropriately added includingfusion detection technology such as“radar+video”,tunnel situation awareness technology and intelligent spherical camera,which canimprove the detection efficiency of abnormal events,and can grasp the real-time operation status of the tunnel and the intelligent focus view afterwards.To effectively solve the problem of relying on manual inspection video monitoring tunnel traffic events,its intelligence,timeliness and accuracy haveplayed an important role in preventing the occurrence of secondary accidents and controlling the development of the situation.