首页|基于多源遥感的铁路外部环境隐患监测方法综述

基于多源遥感的铁路外部环境隐患监测方法综述

扫码查看
随着中国铁路运营里程的不断增加,铁路外部环境日益复杂,以彩钢瓦房、防尘网、塑料大棚、地膜、塑料垃圾为主的铁路外部环境隐患层出不穷,频繁引发铁路交通重大事故,已经成为铁路安全运营的重要制约因素。铁路外部环境隐患的高效监测是实现治理的重要前提,而遥感技术是目前实现大范围、低成本铁路外部环境隐患监测的最佳手段。为此,文章对光学遥感、合成孔径雷达、激光雷达、地基视频监控等 4 类遥感监测技术的优势、局限及应用现状进行了梳理;分析了基于人工特征和基于深度学习这两类监测识别方法的特点及不足;最后,从铁路外部环境多源遥感数据多层次融合、铁路外部环境隐患精准识别模型构建、知识引导的铁路外部环境隐患智能动态监测等方面,对基于多源遥感数据融合的铁路外部环境隐患智能监测进行了展望。
A Review of Multi-Source Remote Sensing Methods for Railway External Environmental Hazards Monitoring
With the continuous increase in the operational mileage of railways in our country,the external environment of railways is becoming increasingly complex.Hazards in the external environment of railways,mainly consisting of color steel tile houses,dust-proof nets,plastic greenhouses,ground films,and plastic waste,emerge frequently.These hazards,which frequently lead to major railway traffic accidents,have become a significant constraint on the safe operation of railways.The efficient monitoring of hazards in the external environment of railways is a crucial prerequisite for governance,and remote sensing technology is currently the optimal means to achieve large-scale,low-cost monitoring of external environmental hazards in railways.In this regard,this paper systematically reviews four categories of remote sensing monitoring technologies:optical remote sensing,synthetic aperture radar,lidar,and ground-based video surveillance.The paper discusses the advantages,limitations,and current applications of these technologies.It analyzes two types of monitoring and identification methods:those based on artificial features and those based on deep learning,highlighting their characteristics and shortcomings.Finally,the paper looks forward to the intelligent monitoring of railway external environmental hazards based on the fusion of multi-source remote sensing data at multiple levels,the construction of precise identification models for railway external environmental hazards,and knowledge-guided intelligent dynamic monitoring of railway external environmental hazards.

multi-source remote sensingmonitoring technologydeep learningknowledge graph

李治泓、朱庆、廖成、胡翰、陈琳

展开 >

西南交通大学地球科学与环境工程学院,成都 611756

多源遥感 监测技术 深度学习 知识图谱

2024

航天返回与遥感
中国航天科技集团公司第五研究院第508研究所

航天返回与遥感

CSTPCD北大核心
影响因子:0.669
ISSN:1009-8518
年,卷(期):2024.45(1)
  • 90