首页|基于点云的建筑物轮廓提取方法

基于点云的建筑物轮廓提取方法

扫码查看
研究如何快速、准确提取出建筑物信息,对于城市建设、发展与管理具有重要意义,本文基于三维(3D)点云数据,研究并实现了建筑物几何信息的提取.本文的建筑物信息提取思路为:首先,为降低点云数据量,提升后续算法运算速率,对坡度滤波算法进行改进并将改进算法用于点云中地面点、非地面点分类,提取得到非地面点;其次,提出改进3D霍夫变换(HT)提取建筑物立面点;最后,基于降维边界索引方法确定建筑物轮廓线并提取得到局部特征.以实测三维点云数据进行实验,结果表明,在复杂的点云空间场景中,本文提出的方法能够有效避免噪声点等的干扰,高效、准确提取得到建筑物轮廓与局部特征,可为智慧城市建设、城市规划建设等领域提供积极的技术支持.
Building contour extraction method based on point cloud
Studying how to quickly and accurately extract building information is of great significance for urban construction,development,and management. Based on three-dimensional (3D) point cloud data,this paper studied and implemented the extraction of geometric information from buildings. The idea of building information extraction in this paper was as follows:firstly,to reduce the amount of point cloud data and improve the subsequent algorithm operation speed,the slope filtering algorithm was improved and applied to the classification of ground and non-ground points in the point cloud,so as to extract non-ground points. Secondly,an improved 3D Hough transform (HT) was proposed to extract building elevation points. Finally,a dimensionality reduction-based boundary index method was used to determine the building contour lines and extract local features. The experimental results using measured 3D point cloud data show that in complex point cloud spatial scenes,the proposed method can effectively avoid interference such as noise points,efficiently and accurately extract building contours and local features,and provide positive technical support for smart city construction,urban planning and construction,and other fields.

three-dimensional (3D) point cloudpoint cloud filteringbuilding contourlocal featuresimproved 3D Hough transform algorithm

刘辉

展开 >

佛山市禅城区测绘成果质量检验中心,广东佛山 528000

三维点云 点云滤波 建筑物轮廓 局部特征 改进3D霍夫变换算法

广东省重点领域研发计划广东省城乡建设绿色发展资金

2020B0101130009粤财建[2022]80号

2024

北京测绘
北京市测绘设计研究院,北京测绘学会

北京测绘

影响因子:0.55
ISSN:1007-3000
年,卷(期):2024.38(8)