机载LiDAR输电线对地距离智能检测  

Intelligent Detection of Distance below Transmission Line Based on Airborne Point Cloud

作　　者：麻卫峰 吴小东 王冲 闻平 曹磊 朱琪 Ma Weifeng;Wu Xiaodong;Wang Chong;Wen Ping;Cao Lei;Zhu Qi(Faculty of Geography,Yunnan Normal University,Kunming650500,Yunnan,China;Power China Kunming Engineering Corporation Limited,Kunming650000,Yunnan,China;Key Laboratory of Resources and Environmental Remote Sensing for Universities in Yunnan,Kunming650500,Yunnan,China;Institute of International Rivers and Eco-security,Yunnan University,Kunming650500,Yunnan,China)

机构地区：[1]云南师范大学地理学部,云南昆明650500 [2]中国电建集团昆明勘测设计研究院有限公司,云南昆明650000 [3]云南省高校资源与环境遥感重点实验室,云南昆明650500 [4]云南大学国际河流与生态安全研究院,云南昆明650500

出　　处：《应用激光》2024年第4期104-112,共9页Applied Laser

基　　金：云南省基础研究计划项目(202401AT070111);云南省博士后定向资助项目(KD-BSH009);云南省教育厅科学研究基金项目(2021J0438);云南师范大学博士启动项目(2022ZB038)。

摘　　要：对地距离是输电线路巡检和电网风险评估的重要指标,受限于复杂线路环境及地形条件,传统人工测量方法存在作业困难、效率低下等不足。以无人机载LiDAR激光点云数据为基础,提出一套高效的输电线路对地距离智能检测技术流程,包括点云分类、最下层输电线提取、对地距离检测及风险点识别等。主要贡献点包括:(1)通过把输电线三维点云投影至二维残差空间,将输电线点云分割过程转换为投影点聚类问题,淡化了数据缺失、点云噪声因素对单根输电线精细分割的消极影响;(2)设计了基于类簇重心分析的最下层输电线识别与提取策略;(3)实现了基于激光点云的输电线对地距离智能化检测。试验结果表明:该方法可快速从原始点云数据中精确识别最下层输电线,并实现对地距离高效检测及风险点精准识别,具有操作简单、自动化高的特点。研究为机载LiDAR输电线路安全巡检提供理论和实践参考。The distance from power lines to ground is an important indicator for transmission line inspection and power grid risk assessment.However,traditional manual measurement methods are difficult to operate and inefficient due to complex environmental and terrain conditions.Based on UAV LiDAR point cloud data,an efficient intelligent detection technology process is proposed,which includes point cloud classification,bottom-layer transmission line extraction,distance detection to the ground,and risk point identification.The main innovations are:(1)By projecting the three-dimensional point cloud of the transmission line into the two-dimensional residual space,the point cloud segmentation process is transformed into a projection point clustering problem,which dilutes the negative impact of data missing and point cloud noise factors on the fine segmentation of a single transmission line;(2)An identification and extraction strategy for the lowest transmission line is designed based on cluster barycentric analysis;(3)The intelligent detection of transmission line to ground distance is realized based on laser point cloud.The experimental results show that this method can quickly and accurately identify the lowest transmission line from the original point cloud data,and realize the efficient detection of ground distance and accurate identification of risk points.It has the characteristics of simple operation and high automation.This research provides a theoretical and practical reference for safety inspection of airborne LiDAR transmission lines.

关 键 词：激光雷达 输电线路 对地距离 智能检测 

分 类 号：TN249[电子电信—物理电子学]