摘要
近年来高速发展的激光雷达技术为滑坡监测提供了新的思路,DEM比较法将格网内的点归算至同一高程面,导致获取的滑坡量不够准确。为解决这一问题,提出一种基于二重积分的滑坡量计算方法。首先利用ICP算法对不同时期的点云进行配准,然后将滑坡区域体素化,以体素为单元,通过主成分分析方法计算各体素内不同测期点云的法向量和拟合平面,最后利用拟合函数通过二重积分计算各体素内的体积变化量。以两期滑坡点云为例进行实验分析,分别采用0.5、1、1.5、2 m分辨率进行测试分析,实验结果表明,所提方法相比于DEM比较法计算的误差在上述分辨率情况下分别减少了150、153、69、142 m^(3),为同类应用提供了理论基础和应用参考。
The LiDAR technology has been developed rapidly in recent years,which provides new ideas for landslide monitoring.The use of DEM comparison method to adjust the points within the grid to the same elevation may result in less accurate measurement of landslide volume.To address this issue,we proposed a landslide volume calculation method using double integration.Firstly,we used ICP algorithm to register the point clouds between different epochs.And then,the landslide area was voxelized using predefined voxel size.Taking the voxel as the unit,we used PCA to estimate the normal vector and the fitted plane of each voxel.Finally,we calculated the volume change within each voxel between two epochs through double integration.We used two sets of landslide point clouds with resolutions of 0.5 m,1 m,1.5 m and 2 m to conduct experimental analysis,respectively.The results show that compared with the DEM comparison method,the landslide volume calculation errors of the proposed method are reduced by 150 m^(3),153 m^(3),69 m^(3)and 142 m^(3)under the above resolutions,respectively,providing a theoretical basis and application reference for similar applications.
作者
王宇翔
沈月千
臧光勇
杨旭
WANG Yuxiang;SHEN Yueqian;ZANG Guangyong;YANG Xu(Jiangsu Geological Engineering Environment Intelligent Monitoring Engineering Research Center,Nanjing 210041,China;School of Earth Sciences and Engineering,Hohai University,Nanjing 211100,China)
出处
《地理空间信息》
2024年第10期91-95,共5页
Geospatial Information
基金
国家自然科学基金青年基金资助项目(41801379)
江苏省地质工程环境智能监控工程研究中心开放基金资助项目(2021-ZNJKJJ-06)。
