基于激光雷达数据的三维建筑屋顶重构算法  

Three-dimensional Building Roof Construction based on LiDAR Data

作　　者：翟月[1] ZHAI Yue(College of Software,Anhui Vocational College of Electronics&Information Technology,Bengbu 233000,China)

机构地区：[1]安徽电子信息职业技术学院软件学院,安徽蚌埠233000

出　　处：《信阳农林学院学报》2022年第1期123-127,132,共6页Journal of Xinyang Agriculture and Forestry University

基　　金：2019年安徽省质量工程教学资源库建设项目(2019zyk31,2019zyk32);2020年安徽省教学示范课项目(2020SJJXSFK0213);2021年安徽电子信息职业技术学院校级项目(2021AHDZZK06)。

摘　　要：针对高精度建筑虚拟化中的建筑屋顶重构问题,提出了基于激光雷达数据驱动的三维建筑屋顶重构算法。首先,根据高度相似性和平面相似性将建筑物标记的点云聚类为同构簇;其次,基于分割的簇,提取线性建模线索,使用二元空间分割技术重构模型线索的拓扑元素;再次,结合最小描述长度框架和假设检验来实现正则化,发掘建筑屋顶模型的规律;最后,使用基准数据集来评估该算法。实验结果表明,所提出的方法可以生成准确的三维建筑屋顶模型。For the problem of building roof reconstruction in high-precision building virtualization,this paper proposes a 3D building roof reconstruction algorithm based on light detection and ranging data.The algorithm first clusters the point clouds of building markers into isomorphic clusters based on high similarity and plane similarity.Next,based on the segmented clusters,linear modeling cues are extracted,and the topological elements of the model cues are reconstructed using binary space segmentation technology.Then,combining the minimum description length framework with the hypothesis and test to perform regularization process,the rules of the building roof model are explored.Finally,it was tested on the benchmark data set.The simulation results show that the proposed method can generate accurate 3D building roof models.

关 键 词：屋顶重构 激光雷达数据 点云 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]