基于多元入侵模式的铁路入侵智能检测方法研究  

Research on intelligent detection method of railway intrusion based on multiple intrusion mode

作　　者：张省伟[1] 王博[1] ZHANG Xing-wei;WANG Bo(Dept. of Traction Power,Xi’an Railway Vocational & Technical Institute,Xi’an 710026,China)

机构地区：[1]西安铁路职业技术学院牵引动力学院

出　　处：《电子设计工程》2019年第17期89-92,97,共5页Electronic Design Engineering

摘　　要：保证铁路行车安全的重要手段之一就是防止铁路入侵行为的发生,在铁路沿线采用视频监控的方式则是预防铁路入侵的有效措施,但由于人工监控可靠性较差,因此采用基于计算机智能算法的铁路入侵检测系统就尤为必要。本文提出了基于多元入侵模式的铁路入侵智能检测方法,即一种对多目标和多组合行为铁路入侵行为建模的数学模型和基于方向场的铁轨识别模型,以及铁路入侵的动态目标检测算法流程。经轨道监控视频实验验证,提出的方法不仅可有效识别视频帧画面中的轨道,还可识别不同入侵目标并根据入侵对象的不同做出跟踪和预警,该方法对实现智能铁路入侵检测具有重要意义。One of the important means to ensure the safety of railway traffic is to prevent the occurrence of railway intrusion. The use of video surveillance along the railway is an effective measure to prevent railway intrusion. However,due to the poor reliability of manual monitoring,computer intelligence based algorithm is used. Railway intrusion detection systems are especially necessary. This paper proposes a railway intrusion intelligence detection method based on multi-intrusion mode,which is a mathematical model for modeling multi-objective and multi-combination behavior railway intrusion behavior and a rail-track identification model based on directional field,and a dynamic target detection algorithm flow for railway intrusion.. The orbit monitoring video experiment proves that the proposed method can not only effectively identify the orbits in the video frame picture,but also identify different intrusion targets and make tracking and early warning according to different intrusion objects. This method is of great significance for realizing intelligent railway intrusion detection.

关 键 词：铁路入侵 智能检测 轨道 安全 

分 类 号：U298.1[交通运输工程—交通运输规划与管理] TP391.4[交通运输工程—道路与铁道工程]