基于部分重合分段的电力线激光点云数据聚类方法与实现

Power Line Laser Point Cloud Data Clustering Algorithm and Implementation Based on Partial Coincidence Segmentation

在无人机激光雷达电力线巡检中,基于激光点云数据的电力线三维重建是首要工作。针对已有方法无法解决多种构型电力线激光点云聚类等问题,提出了基于部分重合分段的电力线激光点云数据聚类方法。该算法包含构建单档单根电力线的三维重建模型中直线段模型和悬链线段模型的数学表达式、数据预处理和部分重合数据分段、模型拟合及线段标号与合成等三个步骤。通过实验验证了该算法的实用性和有效性,实现了对任意长度、任意电塔高差、存在不规则断裂和粗差的电力线点云的正确分割,具备较高精度和鲁棒性。

In the inspection of power line scanned by unmanned aerial vehicle(UAV)with lidar,the three-dimensional(3D)reconstruction of power line based on laser point cloud data is the first task.In order to solve the problem that the existing methods cannot meet the requirements of power line laser point cloud clustering with multiple configurations,a power line laser point cloud data clustering method based on partial overlap segmentation is proposed.The algorithm consists of three steps:constructing mathematical expressions for the linear segment model and the suspended chain line segment model in the 3D reconstructed model of a single-file single power line,data preprocessing and partial overlap data segmentation,model fitting and line segment labeling and synthesis.The practicality and effectiveness of the algorithm are verified through experiments.The algorithm realizes the correct segmentation of power line point clouds of arbitrary length,arbitrary tower height difference,irregular fracture and coarse difference,and has high accuracy and robustness.