分层随机抽样的单档电力线LiDAR点云聚类方法

Clustering of airborne LiDAR point cloud of 3D powerline reconstruction in a span using stratified sampling

针对基于直升机激光雷达(LiDAR)测量技术的电力巡线对全自动电力线三维重建的迫切需求,该文提出了一种基于分层随机抽样和电力线三维重建数学模型约束的单档电力线LiDAR点云聚类方法。其中,利用了直线段和悬链线段相结合的电力线三维数学模型描述了现实三维场景中的电力线形态,并以此约束点云分层随机抽样的结果。聚类过程中,首先进行数据预处理和分段组织;接着进行标号。实验表明,该文的聚类方法具有对电力线根数、电力线类型、电力线空间构型、长度、两端电塔高差、点云不规则断裂、粗差等因素不敏感的优势。

3D powerline reconstruction is one of the main contents in powerline patrols using LiDAR systems mounted on helicopters, and its key step is performing clustering of the powerline LiDAR points of a span. A method was proposed to cluster the powerline LiDAR points in a span based on the stratified sampling technique with the constraint of 3D powerline model. The 3D powerline model was composed of a line segment and a catenary segment, which was a mathematical description of a real powerline shape and was employed to verify the correctness of the sampled points using stratified sampling. In the clustering process, the input data was firstly preprocessed and separated into several parts, and the labeling was secondly performed. The experiments demonstrated that the proposed method was robust to many factors such as the number of the power lines, the types of the power lines, the arrangement of the power lines, the length of the spans, the elevation differences of the pylons, the irregular breakage of the LiDAR point clouds, and the blunders etc.