基于无人机多模态数据的铁路防洪隐患排查系统研发  

Railway Flood Hazard Investigation System Based on Multi-Modal Data from Unmanned Aerial Vehicles

作　　者：陈昱行 高至飞 胡朝鹏 宋国策 CHEN Yuhang;GAO Zhifei;HU Zhaopeng;SONG Guoce(China Railway Design Corporation,Tianjin 300308,China;Guangzhou Railway Science and Technology Development Co.,Ltd.,Guangzhou 510030,China)

机构地区：[1]中国铁路设计集团有限公司,天津300308 [2]广州铁路科技开发有限公司,广州510030

出　　处：《铁道勘察》2024年第5期156-162,共7页Railway Investigation and Surveying

基　　金：中国铁路设计集团有限公司科技开发计划重点课题(2022A02401005)。

摘　　要：随着铁路防洪安全巡检的常态化,传统的现场踏勘方法已无法满足日益增长的防洪隐患排查需求。旨在开发一种集成化的桌面端系统,以提高防洪隐患监测的效率和准确性。通过无人机多模态数据的集成管理和铁路基础地理信息的融合,解决了隐患定位、概况管理和展示汇报的问题;基于WPF框架,融合Cefsharp和Vue.js技术,可实现三维场景与数据操作的紧密结合;利用OpenGL的3D GIS引擎,可提供实景三维模型、倾斜摄影原片、航飞视频等多模态数据的一体化展示与空间分析能力;系统设计分类的隐患标注功能,并通过离线的Spatialite数据库实现数据的持久化和导入导出。选取总里程300 km,包含148个防洪重点工点,2期无人机航飞数据的防洪隐患排查项目作为应用案例,验证了系统的有效性。研究结果表明,该系统能够通过关系数据库和对象存储的结合,实现无人机多模态数据的集成式管理,形成基于高精度实景三维场景的防洪隐患排查机制。The conventional on-site survey for railway flood safety is inadequate for the increasing demand of flood hazard assessment.This study introduced an integrated desktop system to enhance the efficiency and precision of flood hazard monitoring.By integrating drone multimodal data with railway geographic information,the system addressed the challenges of hazard identification,management,and reporting.Based on the WPF framework and integrated with the Cefsharp and Vue.js,it seamlessly combined 3D scenes with data operations and employed an OpenGL-based 3D GIS engine for multimodal data display and spatial analysis.A classified hazard marking function and an offline SpatiaLite database ensured data persistence and portability.Tested on a 300 km railway section with 148 key flood points and two phases of drone data,the system demonstrates effective multimodal data management and real scene visualization for hazard assessment.

关 键 词：铁路 防洪 无人机 多模态数据 地理信息系统 

分 类 号：U212[交通运输工程—道路与铁道工程] P231[天文地球—摄影测量与遥感]