基于高程投影的输电线路激光点云分类研究

Laser point cloud classification of transmission line based on elevation projection

针对传统输电线路激光点云自动分类方法存在分类准确率低,耗时较长、电力巡检质量低等问题,文中提出基于高程投影的输电线路激光点云自动分类方法研究。通过对高程投影法工作原理进行分析,对输电线路激光点云数据进行提取;通过确定塔杆间地物点范围,确定输电线路点云数据提取范围;选择空间距离抽稀的方法进行点云抽稀,将点云密度均一化;利用最小二乘法重建电力线的三维空间,进而实现输电线路的安全检测。实验结果表明:采用文中方法进行输电线路激光点云分类的效率较高,且准确性高达90%以上,具有一定的可行性。

In view of the problems of low classification accuracy,long time-consuming and low quality of power inspection,an automatic classification method for transmission line laser point cloud is proposed based on elevation projection.By analyzing the working principle of elevation projection method,the laser point cloud data of transmission line is extracted,the maximum elevation value and minimum elevation value of transmission line point cloud data are determined,the optimum threshold value of point cloud data is obtained,the range of point cloud data extraction is determined by determining the range of ground object points between tower poles,the spatial distance extraction method is selected to dilute the point cloud density,and the point cloud density is homogenized by multi-sided function fitting,effectively classifying the point cloud data of transmission lines.The least square method is used to reconstruct the three-dimensional space of power lines,and the safety monitoring of transmission lines is realized.The experimental results show that the method is more efficient and more accurate than 90%,which is feasible.