艾海滨,董友强,王庆栋,张力

• 1:中国测绘科学研究院
• 2:中国矿业大学(北京)地球科学与测绘工程学院

摘要(Abstract)：

针对应用倾斜航空影像进行自动化建模所创建的建筑物建模存在诸如模型细节缺失,难以获取语义信息的问题,该文提出了一种采用基于语义的建模方式构建建筑物模型的方法;首先对密集图像匹配(DIM)点云进行建筑物类提取,实现场景中建筑和其它地物的分离;然后进行基于点云分割和建筑物验证的建筑物单体标注,以保证对单体建筑物的有效标记;最后利用倾斜航空影像和DIM点云的建筑物各组件的语义识别和提取算法,实现以单体建筑物为建模对象的模型重建;同时,设计适用于建筑物重建的语义建模框架,对单体语义模型进行自动化纹理映射,构建出语义化的建筑物模型。该文提出的建模关键技术方法将有助于减少城市建模成本,提高建模效率,为智慧城市基础数据的建设提供关键技术储备,并发挥相应的的社会和经济效益。

关键词(KeyWords)： 倾斜摄影;三维建模;语义建模;点云分类;影像匹配点云

Abstract:

Keywords:

基金项目(Foundation):

作者(Author): 艾海滨,董友强,王庆栋,张力

DOI: 10.16251/j.cnki.1009-2307.2019.06.022

参考文献(References)：

• [1]RAU J Y,JHAN J P,HSU Y C.Analysis of oblique aerial images for land cover and point cloud classification in an urban environment[J].IEEE Transactions on Geoscience&Remote Sensing,2015,53(3):1304-1319.
• [2]ANTONARAKIS A S,RICHARDS K S,BRASING-TON J.Object-based land cover classification using airborne LiDAR[J].Remote Sensing of Environment,2008,112:2988-2998.
• [3]ROTTENSTEINER F,SOHN G,GERKE M,et al.Results of the ISPRS benchmark on urban object detection and 3Dbuilding reconstruction[J].ISPRS Journal of Photogrammetry and Remote Sensing,2014,93:256-271.
• [4]王庆栋,艾海滨,张力.利用倾斜摄影和3ds Max技术快速实现城市建模[J].测绘科学,2014,39(6):74-78.(WANG Qingdong,AI Haibin,ZHANG Li.Rapid city modeling based on oblique photogrammetry and 3ds MAX technique[J].Science of Surveying and Mapping,2014,39(6):74-78.)
• [5]张瑞瑞,艾海滨,张力,等.附加约束条件的半自动复杂建筑物重建方法[J].测绘通报,2015(11):64-68.(ZHANG Ruirui,AI Haibin,ZHANG Li,et al.A novel method for semi-automatic complex building reconstruction under additional conditional constraints[J].Bulletin of Surveying and Mapping,2015(11):64-68.)
• [6]詹总谦,林元培,艾海滨.基于3ds Max二次开发的建筑物快速三维重建[J].测绘通报,2016(11):22-25,42.(ZHAN Zongqian,LIN Yuanpei,AI Haibi.Rapid3Dreconstruction based on secondary development of3ds Max[J].Bulletin of Surveying and Mapping,2016(11):22-25,42.)
• [7]ZHANG Jianqing,ZHANG Yong,ZHANG Zuxun.Determination of exterior parameters for video image sequences from helicopter by block adjustment combined vertical and oblique images[J].Proceedings of SPIE:the International Society for Optical Engineering,2003,5286:191-194.
• [8]HIRSCHMULLER H.Accurate and efficient stereo processing by semi-global matching and mutual information[C]∥Proceedings of the 2005IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Washington:IEEE Computer Society,2005:807-814.
• [9]GERKE M.Dense matching in high resolution oblique airborne images[C/OL]∥CMRT09,2009[2019-02-28].https:∥www.isprs.org/proceedings/XXXVIII/3-W4/pub/CMRT09_Gerke.pdf.
• [10]FROMMHOLZ D,LINKIEWICZ M,MEISSNER H,et al.Extracting semantically annotated 3D building models with textures from oblique aerial imagery[C/OL]∥Proceedings of the International Archives of the Photogrammetry,Remote Sensing and Spatial Information Sciences,2015[2019-02-28].https:∥core.ac.uk/download/pdf/31016957.pdf.
• [11]AXELSSON P.Processing of laser scanner data:algorithms and applications[J].ISPRS Journal of Photogrammetry and Remote Sensing,1999,54(2):138-147.
• [12]DONG Youqiang,CUI Ximin,ZHANG Li,et al.An improved progressive TIN densification filtering method considering the density and standard variance of point clouds[J].ISPRS International Journal of GeoInformation,2018,7(10):409.
• [13]LU Xiaohu,YAO Jian,LI Kai,et al.CannyLines:aparameter-free line segment detector[C]∥Proceedings of the IEEE International Conference on Image Processing.Piscataway:IEEE,2015:507-511.
• [14]GRIGILLO D,KANJIR U.Urban object extraction from digital surface model and digital aerial images[J].ISPRS Annals of Photogrammetry,Remote Sensing and Spatial Information Sciences,2012,I-3:215-220.
• [15]HABIB A F,KIM E M,KIM C J.New methodologies for true orthophoto generation[J].Photogrammetric Engineering&Remote Sensing,2007,73(1):25-36.
• [16]WU K S,OTOO E,SUZUKI K.Optimizing two-pass connected-component labeling algorithms[J].Pattern Analysis&Applications,2009,12(2):117-135.
• [17]王庆栋.基于语义建模框架的机载LiDAR点云建筑物三维重建技术研究[D].武汉:武汉大学,2017.(WANG Qingdong.A semantic modeling frameworkbased method for 3Dbuilding reconstruction from airborne LiDAR point clouds[D].Wuhan:Wuhan University,2017.)