基于渐进三角网的线路巡检点云地面点快速提取方法
Fast Ground Points Extraction Method in Electric Line Inspection Based on the Progressive Encryption TIN
-
摘要:目的 在利用激光点云数据进行线路巡检过程中，获取的点云数据需要进行分类和提取才能进行更进一步的应用。其中地面点是线路巡检过程中一类重要数据，现有的基于渐进三角网的地面点提取方法能够准确的提取地面点数据，然而基于渐进三角网的地面点提取算法效率较低，无法对数据进行快速高效处理。方法 针对此问题，提出了基于改进渐进伪三角网的线路巡检激光点云地面点快速提取方法。该在分析渐进三角网特征的基础上认为相比于待判断点所在三角网，与待判断点最近邻的三个点组成的平面与待判断点具有更强的相似性特征，因此采用K近邻算法计算与待判断点最近的三个点组成的平面，通过kd-tree的数据组织形式提高了搜索和插入效率。结果 与基于渐进三角网的方法相比文章提出的方法的在保证提取精度的条件下极大的提高了地面点提取效率。结论 实验证明：改进后的方法极大提高了地面点提取效率，能够满足电力巡检过程中海量数据快速处理的需求。Abstract:
Introduction Point cloud data can be gotten during the electric line inspection using the LiDAR technology. The point cloud should be classified if the point cloud need to be used in the advanced applications. And the ground points classification is a very important progress during the progress of the point cloud classification. And the ground points extraction method based on the progressive Encryption TIN can extract the ground points accurate but the efficient of the method is very low. Method To the problem we proposed an imitation progressive encryption TIN method. The method based on the Progressive Encryption TIN method and assumed that the plane of nearest three points was more accurate the extract the ground compared with the triangulation plane. In this method the kd-tree was used to organization the point to improve the efficiency.Result The purposed method improves the efficiency very much compared with the progressive encryption TIN method with the almost same accurate.Conclusion And the purposed method can meet the requirement of the fast and large data process.
下载:
