收稿：2024-11-11，
纸质出版：2026-01-07
移动端阅览
屋顶特征线是建筑物表面关键的结构特征。针对现有屋顶特征线提取方法存在算法复杂、鲁棒性较低和过于依赖屋顶面分割精度等问题，本文提出一种联合边界三角形检测与双向叠加平移的建筑物屋顶特征线提取方法。首先，采用Delaunay三角剖分建立屋顶点云的三角网格模型，并创建等深森林（Equi-deep forest， E-forest）数据结构用于快速存储和查询各三角形的邻域信息；其次，根据E-forest提供的各三角形邻接关系查询三角网格模型中的独立边确定边界点，并将各三角形中心点沿各自的法向量方向向上、向下平移，通过统计平移后每个点球域内有点截面的数量确定屋脊线上的点；最后，采用D-P算法及最小二乘算法生成屋顶边界线与屋脊线，并结合多特征线交点平面坐标与局部原始点高程精细化各特征线端点，完成对屋顶面特征线的提取工作。本文选用Building3D以及Vaihingen数据集中的12组典型屋顶点云进行测试，并且与现有特征线提取方法进行比较。实验结果表明，本文方法所提取特征线的Mean值约为1倍的平均点间距、LSM高于85%。整体方法稳定可靠、鲁棒性强且能够满足绝大多数建筑物屋顶特征线的提取需求。
Roof feature lines are among the most important structural features of building surfaces and are widely used in 3D building model reconstruction. To address the issues in existing roof feature line extraction methods
such as high complexity
low robustness
and reliance on the accuracy of roof surface segmentation
this study proposes a novel method that combines boundary triangle detection with bidirectional superimposed translation. First
this study adopts Delaunay triangulation to establish the triangular mesh model of the roof point clouds
and an Equi-deep forest data structure is created to quickly store and search the neighborhood information of each triangle. Second
according to the adjacency relationship of each triangle provided by the equi-deep forest
the independent edges in the triangular mesh model are queried to determine the boundary points
and the center points of each triangle are translated upward and downward along their respective normal vectors. The intersecting lines of adjacent roof patches after projection are determined by counting the number of point sections in each point sphere domain after translation
and the points on the intersecting lines are translated in the opposite direction of the normal vector to obtain the ridge line points. Finally
the Ramer–Douglas–Peucker algorithm was used to simplify the closed independent edges to generate the boundary lines of the roof components
including the contours and step structures. The least squares fitting algorithm was used to vectorize the ridge points to generate the ridge lines. Combined with the plane coordinates of the intersection points of the multifeature lines and the elevation of the local original points
the endpoints of each feature line were refined
and feature line extraction free from roof surface segmentation was completed. To verify the effectiveness of the proposed algorithm
this study selected 12 groups of building roof point clouds with typical roof structure features from the Building3D and Vaihingen datasets for testing and compared them qualitatively and quantitatively with existing roof feature line extraction methods. Results showed that the average coordinate deviation of the feature line endpoints extracted by the proposed method was approximately one times the average point spacing
and the line similarity measure was higher than 85%
which was generally higher than that of the two comparison methods. In addition
compared with the half-edge data structure
the proposed equi-deep forest data structure demonstrates higher efficiency and a smaller memory footprint for storing triangular mesh models
which can alleviate the difficulty of calling the triangle adjacency relationship to a certain extent. Overall
the proposed method is suitable for extracting roof ridge lines from dense mesh models. It demonstrates stable performance and strong robustness and meets the requirements for extracting roof feature lines in the vast majority of building cases
thereby providing a novel approach for roof feature line extraction.
Bao D . 2017 . Research on Parametric Modeling Method Of Point Cloud Based on the Data Driven Approach . Wuhan : Wuhan University
鲍娣 . 2017 . 基于数据驱动的点云参数化建模方法研究 . 武汉 : 武汉大学
Bazazian D and Parés M E . 2021 . EDC-Net: edge detection capsule network for 3D point clouds . Applied Sciences , 11 ( 4 ): 1833 [ DOI: 10.3390/app11041833 http://dx.doi.org/10.3390/app11041833 ]
Cai Z , Ma H C and Zhang L . 2023 . Extraction of roof feature lines based on geometric constraints from airborne LiDAR data . Remote Sensing , 15 ( 23 ): 5493 [ DOI: 10.3390/rs15235493 http://dx.doi.org/10.3390/rs15235493 ]
Chen H W , Yuan X C , Wu L S and Wang X H . 2019 . Automatic point cloud feature-line extraction algorithm based on curvature-mutation analysis . Optics and Precision Engineering , 27 ( 5 ): 1218 - 1228
陈华伟 , 袁小翠 , 吴禄慎 , 王晓辉 . 2019 . 基于曲率突变分析的点云特征线自动提取 . 光学精密工程 , 27 ( 5 ): 1218 - 1228 [ DOI: 10.3788/OPE.20192705.1218 http://dx.doi.org/10.3788/OPE.20192705.1218 ]
Coiffier G , Basselin J , Ray N and Sokolov D . 2021 . Parametric surface fitting on airborne Lidar point clouds for building reconstruction . Computer-Aided Design , 140 : 103090 [ DOI: 10.1016/j.cad.2021.103090 http://dx.doi.org/10.1016/j.cad.2021.103090 ]
Cui W , Qiang W Y and Chen X L . 2003 . Image dynamic matching for stereo vision based on structural information between lines . Control and Decision , 18 ( 5 ): 633 - 636
崔巍 , 强文义 , 陈兴林 . 2003 . 基于直线间结构信息的立体视觉图像动态匹配方法 . 控制与决策 , 18 ( 5 ): 633 - 636 [ DOI: 10.3321/j.issn:1001-0920.2003.05.030 http://dx.doi.org/10.3321/j.issn:1001-0920.2003.05.030 ]
Dos Santos R C , Galo M and Carrilho A C . 2019 . Extraction of building roof boundaries from LiDAR data using an adaptive alpha-shape algorithm . IEEE Geoscience and Remote Sensing Letters , 16 ( 8 ): 1289 - 1293 [ DOI: 10.1109/LGRS.2019.2894098 http://dx.doi.org/10.1109/LGRS.2019.2894098 ]
Du S J , Zhang Y S , Zou Z R , Xu S H , He X and Chen S Y . 2017 . Automatic building extraction from LiDAR data fusion of point and grid-based features . ISPRS Journal of Photogrammetry and Remote Sensing , 130 : 294 - 307 [ DOI: 10.1016/j.isprsjprs.2017.06.005 http://dx.doi.org/10.1016/j.isprsjprs.2017.06.005 ]
Feng M L , Zhang T G , Li S C , Jin G Q and Xia Y J . 2020 . An improved minimum bounding rectangle algorithm for regularized building boundary extraction from aerial LiDAR point clouds with partial occlusions . International Journal of Remote Sensing , 41 ( 1 ): 300 - 319 [ DOI: 10.1080/01431161.2019.1641245 http://dx.doi.org/10.1080/01431161.2019.1641245 ]
Gao G . 2014 . Methods of Regularized Building Models Reconstruction from Aerial LiDAR Data . Wuhan : Wuhan University
高广 . 2014 . 利用机载LiDAR数据生成规范化建筑物模型的关键技术研究 . 武汉 : 武汉大学
Gong J , Zhu Q , Zhang H W , Li X M and Zhou D B . 2011 . An adaptive control method of LODs for 3D scene based on R-tree index . Acta Geodaetica et Cartographica Sinica , 40 ( 4 ): 531 - 534
龚俊 , 朱庆 , 章汉武 , 李晓明 , 周东波 . 2011 . 基于R树索引的三维场景细节层次自适应控制方法 . 测绘学报 , 40 ( 4 ): 531 - 534
Hong S X , Yuan F , Wang J X and Qi J J . 2020 . Research on buildings boundary regularization algorithm for aerial LiDAR point clouds . Science of Surveying and Mapping , 45 ( 7 ): 100 - 105 , 125
洪绍轩 , 袁枫 , 王竞雪 , 齐吉婧 . 2020 . 机载LiDAR点云建筑物边界线规则化算法研究 . 测绘科学 , 45 ( 7 ): 100- 105 , 125 [ DOI: 10.16251/j.cnki.1009-2307.2020.07.015 http://dx.doi.org/10.16251/j.cnki.1009-2307.2020.07.015 ]
Hui Z Y , Hu H Y , Li N and Li Z X . 2022 . Improved alpha-shapes building profile extraction algorithm . Laser and Optoelectronics Progress , 59 ( 10 ): 1028006
惠振阳 , 胡海瑛 , 李娜 , 李卓宣 . 2022 . 改进的Alpha-shapes建筑物轮廓线提取算法 . 激光与光电子学进展 , 59 ( 10 ): 447 - 455 [ DOI: 10.3788/LOP202259.1028006 http://dx.doi.org/10.3788/LOP202259.1028006 ]
ISPRS . ISPRS Test Project on Urban Classification and 3D Building Reconstruction [EB/OL ] . ( 2024-11-11 )[ 2025-11-24 ] . https://www.isprs.org/resources/datasets/benchmarks/UrbanSemLab/detection-and-reconstruction.aspx https://www.isprs.org/resources/datasets/benchmarks/UrbanSemLab/detection-and-reconstruction.aspx .
Jiang X . 2019 . Building Change Detection Based on Urban Area Of Airborne LiDAR Point Cloud . Chengdu : Journal of Southwest Jiaotong University
蒋星 . 2019 . 基于机载LiDAR点云城市地区建筑物的变化探测 . 成都 : 西南交通大学
Kaplan G , Comert R , Kaplan O , Matci D K and Avdan U . 2022 . Using machine learning to extract building inventory information based on LiDAR data . ISPRS International Journal of Geo-Information , 11 ( 10 ): 517 [ DOI: 10.3390/ijgi11100517 http://dx.doi.org/10.3390/ijgi11100517 ]
Kong Q Z , Liao L J and Yuan C . 2023 . Rapid generation of editable engineering drawings from 3D point cloud reconstruction for large-scale buildings . Journal of Building Engineering , 63 : 105486 [ DOI: 10.1016/j.jobe.2022.105486 http://dx.doi.org/10.1016/j.jobe.2022.105486 ]
Lv Y X , Zhanf Y D , Dong S Y , Yang L , Zhang Z Y , Li Z R and Hu S J . 2022 . A convex hull-based feature descriptor for learning tree species classification from ALS point clouds . IEEE Geoscience and Remote Sensing Letters , 19 : 6501005 [ DOI: 10.1109/LGRS.2021.3055773 http://dx.doi.org/10.1109/LGRS.2021.3055773 ]
Salehi A and Mohammadzadeh A . 2017 . Building roof reconstruction based on residue anomaly analysis and shape descriptors from lidar and optical data . Photogrammetric Engineering and Remote Sensing , 83 ( 4 ): 281 - 291 [ DOI: 10.14358/PERS.83.4.281 http://dx.doi.org/10.14358/PERS.83.4.281 ]
Shen W , Li J , Chen Y H , Deng L and Peng G X . 2008 . Algorithms study of building boundary extraction and normalization based on LiDAR data . National Remote Sensing Bulletin , ( 5 ): 692 - 698
沈蔚 , 李京 , 陈云浩 , 邓磊 ， 彭光雄 . 2008 . 基于LiDAR数据的建筑轮廓线提取及规则化算法研究 . 遥感学报 , ( 5 ): 692 - 698 [ DOI: 10.11834/jrs.20080590 http://dx.doi.org/10.11834/jrs.20080590 ]
Sun S H and Salvaggio C . 2013 . Aerial 3D building detection and modeling from airborne LiDAR point clouds . IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 6 ( 3 ): 1440 - 1449 [ DOI: 10.1109/JSTARS.2013.2251457 http://dx.doi.org/10.1109/JSTARS.2013.2251457 ]
Tseng Y H and Hung H C . 2016 . Extraction of building boundary lines from airborne LiDAR point clouds // The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences . Prague : ISPRS: 957 - 962 [ DOI: 10.5194/isprs-archives-XLI-B3-957-2016 http://dx.doi.org/10.5194/isprs-archives-XLI-B3-957-2016 ]
Vanian V , Zamanakos G and Pratikakis I . 2022 . Improving performance of deep learning models for 3D point cloud semantic segmentation via attention mechanisms . Computers and Graphics , 106 : 277 - 287 [ DOI: 10.1016/j.cag.2022.06.010 http://dx.doi.org/10.1016/j.cag.2022.06.010 ]
Wang J X , Song W D and Wang W X . 2012 . A multi-view image matching algorithm for feature point under the moving Z-Plane constraint . National Remote Sensing Bulletin , 16 ( 4 ): 691 - 704
王竞雪 , 宋伟东 , 王伟玺 . 2012 . 移动高程平面约束的多视影像特征点匹配 . 遥感学报 , 16 ( 4 ): 691 - 704 [ DOI: 10.11834/jrs.20121147 http://dx.doi.org/10.11834/jrs.20121147 ]
Wang J X , Zhu Q , Wang W X and Zhao L K . 2014 . Straight line extraction algorithm by Hough transform combining edge grouping . National Remote Sensing Bulletin , 18 ( 2 ): 378 - 389
王竞雪 , 朱庆 , 王伟玺 , 赵丽科 . 2014 . 结合边缘编组的Hough变换直线提取 . 遥感学报 , 18 ( 2 ): 378 - 389 [ DOI: 10.11834/jrs.20143127 http://dx.doi.org/10.11834/jrs.20143127 ]
Wang J X , Jiang Y and Wang L Q . 2023 . Segmentation of building roofs from airborne LiDAR point clouds using voxel-based region growing . Journal of Geo-information Science , 25 ( 12 ): 2468 - 2486
王竞雪 , 蒋莹 , 王丽芹 . 2023 . 基于体素增长的机载LiDAR点云建筑物屋顶面分割 . 地球信息科学学报 , 25 ( 12 ): 2468 - 2486 [ DOI: 10.12082/dqxxkx.2023.230305 http://dx.doi.org/10.12082/dqxxkx.2023.230305 ]
Wang R S , Huang S F and Yang H X . 2023 . Building3 D: An urban-scale dataset and benchmarks for learning roof structures from point clouds// International Conference on Computer Vision . Paris : ICCV: 20019 - 20029 [ DOI: 10.48550/arXiv.2307.11914 http://dx.doi.org/10.48550/arXiv.2307.11914 ]
Wen X D , Chen W M , Xie H and Yan L . 2019 . A method for building model reconstruction based on multi-source feature fusion . Geomatics and Information Science of Wuhan University , 44 ( 5 ): 731 - 736 , 764
文学东 , 陈为民 , 谢洪 , 闫利 . 2019 . 一种融合多源特征的建筑物三维模型重建方法 . 武汉大学学报(信息科学版) , 44 ( 5 ): 731- 736 , 764 [ DOI: 10.13203/j.whugis20180320 http://dx.doi.org/10.13203/j.whugis20180320 ]
Wu H Y , Li X K and Hu Y . 2007 . Approach on interactive extraction of gable-roofed building models from airborne Lidar data . Journal of Image and Graphics , 12 ( 3 ): 474 - 481
吴华意 , 李新科 , 胡勇 . 2007 . 从机载激光扫描数据中交互式提取人字形房屋模型的方法研究 . 中国图象图形学报 , 12 ( 3 ): 474 - 481 [ DOI: 10.3969/j.issn.1006-8961.2007.03.016 http://dx.doi.org/10.3969/j.issn.1006-8961.2007.03.016 ]
Wu Y , Wang L Y , Hu C X and Cheng L . 2021 . Extraction of building contours from airborne LiDAR point cloud using variable radius Alpha Shapes method . Journal of Image and Graphics , 26 ( 4 ): 910 - 923
伍阳 , 王丽妍 , 胡春霞 , 程亮 . 2021 . 可变半径Alpha Shapes提取机载LiDAR点云建筑物轮廓 . 中国图象图形学报 , 26 ( 4 ): 910 - 923 [ DOI: 10.11834/jig.200073 http://dx.doi.org/10.11834/jig.200073 ]
Yang B S , Huang R G , Li J P , Tian M , Dai W X and Zhong R F . 2017 . Automated reconstruction of building LoDs from airborne LiDAR point clouds using an improved morphological scale space . Remote Sensing , 9 : 14 [ DOI: 10.3390/rs9010014 http://dx.doi.org/10.3390/rs9010014 ]
Yang J T , Kang Z Z and Akwensi P H . 2021 . A label-constraint building roof detection method from airborne LiDAR point clouds . IEEE Geoscience and Remote Sensing Letters , 18 ( 8 ): 1466 - 1470 [ DOI: 10.1109/LGRS.2020.2999818 http://dx.doi.org/10.1109/LGRS.2020.2999818 ]
Yang L J , CuiY L , Yang Z Q , Zhai G J and Wang C . 2023 . Multi-level index structure based on spatial distribution characteristics of point cloud . Laser and Infrared , 53 ( 1 ): 137 - 145
杨丽娟 , 崔钰琳 , 杨紫骞 , 翟光杰 , 王超 . 2023 . 基于点云空间分布特征的多级索引结构 . 激光与红外 , 53 ( 1 ): 137 - 145 [ DOI: 10.3969/j.issn.1001-5078.2023.01.021 http://dx.doi.org/10.3969/j.issn.1001-5078.2023.01.021 ]
Yu J Z , Wang J X , Zang D D and Xie X . 2024 . A feature line extraction method for building roof point clouds considering the grid center of gravity distribution . Remote Sensing , 16 ( 16 ): 2969 [ DOI: 10.3390/rs16162969 http://dx.doi.org/10.3390/rs16162969 ]
Zhang R , Li G Y , Wang L , Li M L and Zhou Y L . 2018 . A new method of hybrid index for mobile LiDAR point cloud data . Geomatics and Information Science of Wuhan University , 43 ( 7 ): 993 - 999
张蕊 , 李广云 , 王力 , 李明磊 , 周阳林 . 2018 . 车载LiDAR点云混合索引新方法 . 武汉大学学报(信息科学版) , 43 ( 7 ): 993 - 999 [ DOI: 10.13203/j.whugis20160441 http://dx.doi.org/10.13203/j.whugis20160441 ]
Zhang Y , Liu Z , Liu T P , Peng B , Li X and Zhang Q Y . 2020 . Large-scale point cloud contour extraction via 3-D-guided multiconditional residual generative adversarial network . IEEE Geoscience and Remote Sensing Letters , 17 ( 1 ): 142 - 146 [ DOI: 10.1109/LGRS.2019.2917319 http://dx.doi.org/10.1109/LGRS.2019.2917319 ]
Zhang Y H , Geng G H and Wei X R . 2015 . Valley-ridge feature extraction from point clouds . Optics and Precision Engineering , 23 ( 1 ): 310 - 318
张雨禾 , 耿国华 , 魏潇然 . 2015 . 散乱点云谷脊特征提取 . 光学 精密工程 , 23 ( 1 ): 310 - 318 [ DOI: 10.3788/OPE.20152301.0310 http://dx.doi.org/10.3788/OPE.20152301.0310 ]
Zhao Z . 2020 . Research on 3D Building Model Reconstruction from Airborne LiDAR Point Cloud Aided by Aerial Image . Zhengzhou : PLA Strategic Support Force Information Engineering University
赵传 . 2020 . 航空影像辅助机载LiDAR点云的建筑物三维模型重建技术研究 . 郑州 : 战略支援部队信息工程大学
Zhu Q , Yang Y B and Zhang N N . 2018 . Roof edge extraction and optimization based on aerial image and LiDAR . Geospatial Information , 16 ( 4 ): 36 - 39
朱琴 , 杨英宝 , 张宁宁 . 2018 . 利用LiDAR和航空影像的屋顶边缘提取及优化 . 地理空间信息 , 16 ( 4 ): 36 - 39 [ DOI: 10.3969/j.issn.1672-4623.2018.04.012 http://dx.doi.org/10.3969/j.issn.1672-4623.2018.04.012 ]
相关文章
相关作者
相关机构
京公网安备11010802024621
