输电线路螺栓检测的激光点云数据自动化配准研究
Research on Automatic Registration of Laser Point Cloud Data for Transmission Line Bolt Detection
黄绪勇 1林中爱 1唐标 1滕启韬1
作者信息
- 1. 云南电网有限责任公司电力科学研究院,昆明 650012
- 折叠
摘要
输电线路螺栓较小,检测难度较高,易导致螺栓漏检、错误检测,增加故障和安全事故风险.因此,提出一种面向输电线路螺栓检测的激光点云数据自动化配准方法.建立激光点云数据均匀化空间,通过约束格网范围均匀点云数据特征量,完成抽稀化处理;在点云数据集中选取任意目标点,通过调节空间域和频率域高斯核函数大小均衡云数据噪声数值;使用k-means聚类算法对点云数据进行分类;根据点云数据的三维空间坐标,计算目标配准点与源点在不同方向上的坐标值,通过坐标对比完成自动化配准.实验结果表明,所提方法在不同高度、不同距离以及不同方向上的配准准确性较高,且RMSE指标数值最大值为0.3,表明配准值与真实值拟合程度较高,具有较高的配准精准度.
Abstract
The bolts of transmission lines are relatively small and difficult to detect,which can easily lead to bolt leak-age and incorrect detection,increasing the risk of faults and safety accidents.Therefore,a laser point cloud data auto-matic registration method for transmission line bolt detection is proposed.Establish a homogenization space for laser point cloud data,and complete the thinning process by constraining the feature quantity of uniform point cloud data within the grid range.Select any target point in the point cloud dataset,and balance the noise value of the cloud data by adjusting the size of the Gaussian kernel function in the spatial and frequency domains.Using k-means clustering algorithm to classify point cloud data.Based on the three-dimensional spatial coordinates of point cloud data,calculate the coordinate values of the target registration point and the source point in different directions,and complete automat-ed registration through coordinate comparison.The experimental results show that the proposed method has high regis-tration accuracy at different heights,distances,and directions,and the maximum RMSE index value is 0.3,indicating a high degree of fitting between the registration value and the true value,with high registration accuracy.
关键词
输电线路螺栓检测/激光点云数据/抽稀化/三维空间坐标/自动化配准Key words
transmission line bolt inspection/laser point cloud data/dilution/three dimensional spatial coordinates/au-tomated registration引用本文复制引用
基金项目
云南电网有限责任公司电力科学研究院项目(2021-048)
出版年
2024
NETL
NSTL
维普
万方数据