基于SfM方法生成的密集点云数据的典型建筑物分类研究

地震 ›› 2017, Vol. 37 ›› Issue (3): 127-137.

基于SfM方法生成的密集点云数据的典型建筑物分类研究

张雪华, 王晓青, 袁小祥, 王金霞

中国地震局地震预测研究所, 北京 100036

收稿日期:2017-01-04 发布日期:2019-08-09
作者简介:张雪华(1990-), 男, 山东聊城人, 在读硕士研究生, 主要从事遥感、地理信息系统在地震应急及评估方面的应用研究。
基金资助:
中国地震局地震行业专项(201508010)资助

Classification of Typical Building Roof Geometry Types based on Dense Point Clouds of SfM Method

ZHANG Xue-hua,WANG Xiao-qing,YUAN Xiao-xiang,WANG Jin-xia

Institute of Earthquake Science, CEA, Beijing 100036, China

Received:2017-01-04 Published:2019-08-09

摘要/Abstract

摘要： 建筑物类型的研究对于震后救援和损失评估具有重要作用, 目前针对SfM(Structure from Motion)方法生成的三维密集点云数据的建筑物分类研究较少。首先基于SfM原理生成密集点云, 然后通过建筑物单体点云高度均值和标准差对建筑物的高度和屋顶类型进行初步分类, 进一步提出了建筑物单体屋顶最高点与最低点点云中心点的水平距离因子对单坡和双坡屋顶类型进行再分类的方法。以玛曲县城为研究区, 使用SfM算法对无人机影像进行处理, 并利用上述多因子再分类方法进行建筑物高度和类型分类。实验结果表明, 设置高度均值和标准差阈值分别为6 m和0.25 m时能够准确区分单层、非单层建筑物和平、坡屋顶类型建筑物; 对于单坡和双坡顶建筑物, 利用距离因子, 设定距离阈值1.5 m时可完全区分。对该地区典型建筑物的研究结果表明, 通过基于点云分析的建筑物高度和类型提取方法, 可为地震灾害风险分析和未来潜在地震灾害损失预测所需的建筑物信息的提取提供重要参考。

关键词: SfM, 三维点云, 建筑物分类, 地震灾害评估

Abstract: Studies of building types play an important role in the earthquake rescue and damage assessment, but now research of building classification is relatively few based on dense point clouds of SfM (Structure from Motion) method. Firstly, this paper use SfM method to generate dense points, and classify the buildings height and the roof types with the mean and standard deviation of the height of dense point clouds, and then use the factor of the plane distance of the averages of high and low points to classify the roof types. The study area of this paper is Maqu country, and we processed the UAV images using SfM method and used the factors to classify the height and roof types. For the dense point clouds data from SfM method, the experimental results show that when setting the threshold of height and standard deviations of 6 m and 0.25 m respectively can accurately distinguish the layer number of building and the roof. Using the distance factor to distinguish the shed and dual slope construction, when the distance threshold is 1.5 m, we will get ideal result. Classification of typical buildings with dense point clouds can provide a certain reference function for earthquake disaster risk analysis and future earthquake disaster loss assessment.

Key words: Building roof geometry feature, SfM, 3D point clouds, Classifying factors

中图分类号:

P315.7

张雪华, 王晓青, 袁小祥, 王金霞. 基于SfM方法生成的密集点云数据的典型建筑物分类研究[J]. 地震, 2017, 37(3): 127-137.

ZHANG Xue-hua,WANG Xiao-qing,YUAN Xiao-xiang,WANG Jin-xia. Classification of Typical Building Roof Geometry Types based on Dense Point Clouds of SfM Method[J]. EARTHQUAKE, 2017, 37(3): 127-137.

参考文献

[1] 张培震. 中国地震灾害与防震减灾[J]. 地震地质, 2008, 30(3): 577-583.
[2] 王晓青, 黄树松, 丁香, 等. 尼泊尔8.1级地震建筑物震害遥感提取与分析[J]. 震灾防御技术, 2015, 10(3): 481-490.
[3] 王晓青, 魏成阶, 苗崇刚, 等. 震害遥感快速提取研究以2003年2月24日巴楚—伽师6.8级地震为例[J]. 地学前缘, 2003, 10(s1): 285-291.
[4] 朱博勤, 魏成阶, 张渊智. 航空遥感地震灾害信息的快速提取[J]. 自然灾害学报, 1998, 7(1): 34-39.
[5] Trinder J C, Salah M. Aerial images and LIDAR data fusion for disater change detection[J]. Remote Sensing and Spatial Information Sciences, 2012, 1-4: 227-232.
[6] 窦爱霞, 马宗晋, 黄文丽, 等. 基于机载LiDAR和多光谱图像的建筑物震害自动识别方法[J]. 遥感信息, 2013, 28(4): 103-109.
[7] Johnson K, Nissen E, Saripalli S, et al. Rapid mapping of ultrafine fault zone topography with structure from motion[J]. Geosphere, 2014, 5(10): 969-986.
[8] Smith M J, Chandler J, Rose J. High spatial resolution data acquisition for the geosciences: kite aerial photography[J]. Earth Surface Processes and Landforms, 2009, 34(1): 155-161.
[9] Eisenbeiss H, Lambers K, Sauerbier M. Photogrammetric documentation of an archaeological site (Palpa, Peru) using an autonomous model helicopter[J]. Proc Cipa XX International Symposium, 2005, 238-243.
[10] Niethammer U, Rothmund S, James M. UAV-based remote sensing of landslides[J]. International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 2010, ISPRS Comm. V(XXXVIII): Newcastle-upon-Tyne, U.K.
[11] James M R, Robson S. Straightforward reconstruction of 3D surfaces and topography with a camera: Accuracy and geoscience application[J]. Journal of Geophysical Research: Earth Surface, 2012, 117(F3): 94-96.
[12] Lucieer A, Jong S M D, Turner D. Mapping landslide displacements using Structure from Motion (SfM) and image correlation of multi-temporal UAV photography[J]. Progress in Physical Geography, 2014, 38(1): 97-116.
[13] Vosselman G. Building Reconstruction Using Plana r Faces in Very High Density Height Data[C]. International Archives of Photogrammetry and Remote Sensing, 1999, 32(3/2): 87-92.
[14] Morgan M, Habib A. 3D TIN for Automatic Building Extraction from airborne Lase r Scanning Data[C]. The ASPRS Gateway to the New Millennium, St. Louis, Missouri, 2001.
[15] Hofmann A D. Analysis of T in-structure Parameter Spaces in Airborne Laser Scanner Data for 3D Building Model Generation[C]. International Archives of Photogrammetry, Remote Sensing and Spatial Information Science, 2004, 35(B3): 302-307.
[16] 尤红建, 苏林, 李树楷. 利用机载三维成像仪的DSM数据自动提取建筑物[J]. 武汉大学学报信息科学版, 2002, 27(4): 408-413
[17] Awrangjeb M, Fraser C. Automatic segmentation of raw LiDAR data for extraction of building roofs[J]. Remote Sensing, 2014, 6(5): 3716-3751.
[18] Westoby M J, Brasington J, Glasser N F, et al. ‘Structure-from-Motion’ photogrammetry: A low-cost, effective tool for geoscience applications[J]. Geomorphology, 2012, 179: 300-314.
[19] Snavely N. Scene reconstruction and visualization from Internet photo collections: A Survey[J]. Ipsj Transactions on Computer Vision and Applications, 2011, 3: 44-66.
[20] 魏占玉, Arrowsmith Ramon, 何宏林, 等. 基于SfM方法的高密度点云数据生成及精度分析[J]. 地震地质, 2015, 37(02): 636-648.
[21] Mikolajczyk K, Schmid C. Scale & Affine Invariant Interest Point Detectors[J]. International Journal of Computer Vision, 2004, 1(60): 63-86.
[22] Lowe D G. Distinctive Image Features from Scale-Invariant Keypoints[J]. International Journal of Computer Vision, 2004, 60(2): 63-86.
[23] 曾文宪, 陶本藻. 三维坐标转换的非线性模型[J]. 武汉大学学报(信息科学版), 2003, 28(5): 566-568.
[24] 李陈侠. 东昆仑断裂带东段(玛沁—玛曲)晚第四纪长期滑动习性研究[D]. 中国地震局地质研究所, 2009.
[25] Zhang W, Qi J, Wan P, et al. An Easy-to-Use Airborne LiDAR Data Filtering Method Based on Cloth Simulation[J]. Remote Sensing, 2016, 6(8): 501.