测绘通报 ›› 2021, Vol. 0 ›› Issue (10): 67-72,131.doi: 10.13474/j.cnki.11-2246.2021.307

• 学术研究 • 上一篇    下一篇

LiDAR测量点云融合影像的分块滤波方法

毛冬海1,2, 李守军1,2,3, 王锋4, 戴华毅1, 阳凡林1,5   

  1. 1. 山东科技大学测绘与空间信息学院, 山东 青岛 266590;
    2. 自然资源部海底科学重点实验室, 浙江 杭州 310012;
    3. 自然资源部第二海洋研究所, 浙江 杭州 310012;
    4. 浙江省工程物探勘察设计院有限公司, 浙江 杭州 310005;
    5. 自然资源部海洋测绘重点实验室, 山东 青岛 266590
  • 收稿日期:2021-02-01 修回日期:2021-08-14 出版日期:2021-10-25 发布日期:2021-11-13
  • 通讯作者: 李守军。E-mail:0911guang@163.com
  • 作者简介:毛冬海(1996-),男,硕士生,研究方向为无人机航空摄影测量与数据处理方法。E-mail:mdhwd1996@163.com
  • 基金资助:
    国家重点研发计划专项(2016YFC1401210);国家自然科学基金(41576099);全球变化与海气相互作用专项二期(20094G);浙江省沿海滩涂(潮间带)资源监测与评价项目(SJ21003);中央级公益性科研院所基本科研业务费专项(JG2110)

Block filtering method for LiDAR point cloud fusion image

MAO Donghai1,2, LI Shoujun1,2,3, WANG Feng4, DAI Huayi1, YANG Fanlin1,5   

  1. 1. College of Geodesy and Geomatics, Shandong University of Science and Technology, Qingdao 266590, China;
    2. Key Laboratory of Submarine Geosciences, Ministry of Natural Resources, Hangzhou 310012, China;
    3. Second Institute of Oceanography, Ministry of Natural Resources, Hangzhou 310012, China;
    4. Zhejiang Engineering Geophysical Prospecting and Design Institute Co., Ltd., Hangzhou 310005, China;
    5. Key Laboratory of Marine Surveying and Mapping, Ministry of Natural Resources, Qingdao 266590, China
  • Received:2021-02-01 Revised:2021-08-14 Online:2021-10-25 Published:2021-11-13

摘要: 针对现有LiDAR地面点滤波算法对复杂地形地物适应性不强的问题,本文提出了一种融合点云与地面影像分块滤波的方法。首先,将地面影像与点云匹配,使点云从影像中获取更多的光谱纹理信息。然后,分析地物光谱、林地相对密度、点云高程特征、地面DSM模型及其坡度,并基于决策级融合将原始点云切割成若干独立的区块。最后,根据每块区域不同的多元细节特征,对IPTD滤波算法进行改进并利用搜索法优化参数,得到最优且稳健的结果。利用滤波后的总地面点通过插值算法得到的DEM模型和相关试验验证了本文算法的优越性。

关键词: 数据融合, LiDAR点云, 分块滤波, 多元特征, 数字高程模型

Abstract: Aiming at the problem that the existing LiDAR ground point filtering algorithm is not adaptable to complex terrain and objects, a block filtering method combining point cloud and ground image is proposed in this paper. Firstly, the ground image is matched with the point cloud to make it obtain more spectral texture information from the image. Secondly, the ground feature spectrum, forest land relative density, point cloud elevation characteristics, DSM model and its slope are fully analyzed, and the original point cloud is cut into several independent blocks based on decision-making level fusion. Finally, according to the different multivariate detail characteristics of each region, the IPTD filtering algorithm is improved and the parameters are optimized by the search method to obtain the optimal and robust results. Using the filtered total ground points, the DEM model obtained by interpolation algorithm and related experiments verify the superiority of the proposed algorithm.

Key words: data fusion, LiDAR point cloud, segmentation filtering, multivariate characteristics, DEM

中图分类号: