多尺度超体素生长的点云平面分割算法

Point Cloud Plane Segmentation Algorithm for Multi-Scale Supervoxel Growth

ES评分 0

DOI 10.14075/j.jgg.2024.10.462
刊名
Journal of Geodesy and Geodynamics
年,卷(期) 2025, 45(8)
作者
作者单位

1 湖南省遥感地质调查监测所,长沙市劳动西路256号,410015;
2 武汉理工大学安全科学与应急管理学院,武汉市珞狮路205号,430070;
3 武汉大学测绘学院,武汉市珞喻路129号,430079

摘要
为同时提高点云平面分割的效率与可靠性,在基于多尺度超体素及区域生长算法的基础上,以超体素为生长单元,采用体素区域生长算法实现精细点云平面分割。结果表明,本文方法对室内外点云数据都有较好的分割效果,优于RANSAC、Khaloo及全局能量优化方法,且本文方法的Precision、Recall和F1-score都超过0.90。
Abstract
In order to simultaneously improve the efficiency and reliability of point cloud plane segmentation, based on the multi-scale supervoxel and region growing algorithm, supervoxels are used as the growing units, and the voxel region growing algorithm is used to achieve fine point cloud plane segmentation. The results show that the proposed method has good segmentation effect on both indoor and outdoor point cloud data, which is superior to RANSAC, Khaloo and global energy optimization methods. Moreover, the Precision, Recall and F1-score of the method in this paper are all higher than 0.90.
关键词
超体素; 区域生长; 分割; 多尺度; 平面点云
KeyWord
hypervoxel; regional growth; segmentation; multi-scale; plane point cloud
基金项目
页码 813-817
  • 参考文献
  • 相关文献
  • 引用本文

聂道许, 毛娟娟, 花向红, 陈西江. 多尺度超体素生长的点云平面分割算法 [J]. 大地测量与地球动力学. 2025; 45; (8). 813 - 817.

  • 文献评论

相关学者

相关机构