摘要
针对传统二维图像方法在高压电缆接头表面不规则及遮挡条件下存在精度不足的问题,提出一种基于三维点云的电缆接头外径参数提取算法,包括点云姿态归一化、双向切片局部区域构建、曲率加权关键点提取及对称关键点配对外径估计等步骤。在真实接头点云数据集上进行实验验证,结果表明:该方法在完整、遮挡及降采样等不同工况下均表现出较好的鲁棒性与精度,平均测量误差为0.40 mm,处理效率随数据量呈线性变化,能够满足工程应用对精度与效率的双重要求。
Aiming at the problem that the traditional 2D image method has insufficient accuracy under the irregular surface of high-voltage cable joints and occlusion conditions,an algorithm based on 3D point cloud for extracting the outer diameter parameters of cable joints is proposed,including the steps of point cloud pose normalization,two-way slice local area construction,curvature weighted key point extraction and symmetrical key point pairing outer diameter estimation.Experimental validation on real joint point cloud datasets demonstrates that this method exhibits robust performance and high accuracy under various conditions,including complete coverage,occlusion,and downsampling.The average measurement error is 0.40 mm,and processing efficiency scales linearly with data volume.This approach effectively meets the dual requirements of precision and efficiency demanded by engineering applications.
作者
许鹏程
梅云初
吴奕锴
王建军
李天娇
郑钰川
高敬贝
XU Peng-cheng;MEI Yun-chu;WU Yi-kai;WANG Jian-jun;LI Tian-jiao;ZHENG Yu-chuan;GAO Jing-bei(Training Center of State Grid Shanghai Electric Power Company,Shanghai 200438,China)
出处
《电工电气》
2025年第12期53-57,共5页
Electrotechnics Electric
基金
国网上海市电力公司科技项目(B30970240001)。
关键词
三维点云
高压电缆
外径测量
关键点提取
曲率加权
3D point cloud
high-voltage cable
outer diameter measurement
key point extraction
curvature weighted
