测绘通报 ›› 2021, Vol. 0 ›› Issue (8): 33-36.doi: 10.13474/j.cnki.11-2246.2021.0236

• 学术研究 • 上一篇    下一篇

以窗口化和地形坡度为基础的植被茂密区域点云滤波算法

马涛1, 杨小明2, 阎跃观1, 闫东川3   

  1. 1. 中国矿业大学(北京), 北京 100083;
    2. 河北工业大学, 天津 300400;
    3. 中国冶金地质总局矿产资源研究院, 北京 101300
  • 收稿日期:2020-09-08 修回日期:2020-09-28 发布日期:2021-08-30
  • 通讯作者: 闫东川。E-mail:yandongchuan@cmgb.cn
  • 作者简介:马涛(1996-),男,硕士生,研究方向为三维激光扫描。E-mail:hebut_mt@163.com
  • 基金资助:
    国家自然科学基金(51404272)

Point cloud filtering algorithm based on windowing and terrain slope in densely-vegetated area

MA Tao1, YANG Xiaoming2, YAN Yueguan1, YAN Dongchuan3   

  1. 1. China University of Mining and Technology-Beijing, Beijing 100083, China;
    2. Hebei University of Technology, Tianjin 300400, China;
    3. Institute of Mineral Resources, General Administration of Metallurgical Geology of China, Beijing 101300, China
  • Received:2020-09-08 Revised:2020-09-28 Published:2021-08-30

摘要: 三维激光扫描仪获得经典地貌的点云数据,需进行滤波剔除地面植被。由于植被茂密区域点云密集或遮挡,地面点极少,无法拟合出地形表面,这部分植被点很难剔除。针对植被茂密区域点云数据的特点,本文提出以窗口化和地形坡度为基础的植被茂密区域点云滤波算法,认为非地形坡度引起的高程差异的两相邻点中,较高的点为非地面点。试验结果表明,本文算法可以很好地去除植被茂密区域中低矮的植被点,保留真实的地面点,提高了植被茂密区域点云滤波的处理精度。

关键词: 三维激光扫描, 点云处理, 植被茂密区域, 滤波算法, 窗口化, 地形坡度

Abstract: The three-dimensional laser scanner obtains the point cloud data of the classic landform, which needs to be filtered to remove the ground vegetation. Because the point cloud is dense or occluded in densely-vegetated areas, the ground points are very few, and the terrain surface cannot be fitted. This part of the vegetation points is difficult to remove. Aiming at the characteristics of the point cloud data of densely-vegetated areas, this paper proposes a point cloud filtering algorithm for densely-vegetated areas based on windowing and terrain slope. It is considered that the higher point of the two adjacent points of the elevation difference caused by non-terrain slope is non-ground point. The test results show that this algorithm can well remove low vegetation points in densely-vegetated areas, retain real ground points, and improve the processing accuracy of point cloud filtering in densely-vegetated areas.

Key words: 3D laser scanning, point cloud processing, densely-vegetated areas, filtering algorithm, windowing, terrain slope

中图分类号: