一种基于曲面变化度的点云精简算法OA
A point cloud simplification method based on surface variation
三维激光扫描具备高效率、高精度、高分辨率等特点,是空间地理信息数据采集的主要技术手段,但其获取的点云数据量庞大,数据处理对软硬件资源要求高、处理效率低.为提升点云数据处理效率,通常需要对点云进行优化处理,其原则是在保留被测对象主要特征基础上,尽量降低点云密度.本文给出了一种基于曲面变化度的点云精简算法:先计算点云曲面变化度,再根据曲面变化度并结合对象特征将点云划分为"缓变区"和"陡变区",对缓变区采用体素格法精简点云,对陡变区采用曲率精简法.结果表明,该算法能有效保留对象主要特征,并大幅降低点云密度、减少点云数量,可为三维激光扫描数据处理和应用提供参考.
Three-dimensional(3D)laser scanning technology is characterized by high efficiency,high precision,and high resolu-tion,serving as a core technical approach for spatial geographic information data acquisition.However,the point clouds generated by this technology feature massive data volumes,which impose stringent requirements on hardware and software environments during processing and lead to low processing efficiency.To improve the efficiency of point cloud data processing,point clouds are typically subjected to optimization workflows,with the core principle of minimizing point cloud density as much as possible while preserving the key features of the scanned targets.This paper proposed a point cloud simplification algorithm based on surface variation degree.First,the surface variation degree of the input point cloud was calculated.Next,the point cloud was partitioned into gently varying regions and steeply varying regions based on the computed surface variation degree and object feature con-straints.The voxel grid method was adopted for simplification of points in gently varying regions,while the curvature-based sim-plification method was applied to points in steeply varying regions.Experimental results demonstrate that the proposed algorithm can effectively retain the primary features of the scanned objects,significantly reduce point cloud density and the total number of points.It thus provides a valuable reference for the processing and application of 3D laser scanning data.
杨学彬;邓勇;彭云
中国电建集团中南勘测设计研究院有限公司,湖南 长沙 410007中国电建集团中南勘测设计研究院有限公司,湖南 长沙 410007中国电建集团中南勘测设计研究院有限公司,湖南 长沙 410007
天文与地球科学
激光扫描点云优化点云精简体素格法曲率精简法
laser scanningpoint cloud optimizationpoint cloud simplificationvoxel grid methodcurvature simplifica-tion method
《北京测绘》 2026 (8)
1171-1174,4
评论