基于三维激光扫描的BIM逆向建模技术研究热点和趋势分析

Visual Analysis of Research on BIM and 3D Laser Scanning

BIM技术以其三维可视化程度高、信息整合能力强等优势,广泛地应用于建筑全生命周期中。三维激光扫描技术生成的点云数据拥有精度高和非接触式等特点,大量应用于智慧城市、文物保护、基础测绘等领域。基于三维激光扫描的BIM逆向建模技术是BIM技术的较新应用方向。为客观地反映目前三维激光逆向建模技术研究成果,本文利用CiteSpace软件对2011-2020年间收录在Web of Science核心合集中的166篇相关文献进行可视化分析。结果表明:自三维激光逆向建模技术问世后,学者们主要围绕着“模型重建”、“点云分割”、“机器学习”等热点展开研究;该领域的研究文献发表数量逐年增加,但关键作者和机构分布较为松散,凝聚力不够;发文最多的机构分别是Heriot-Watt University和Poltecn Milan,发文最多的作者是Bosche F等,发文量最多的国家是美国和西班牙,中国位居第三。目前,该领域的研究重点已经从ICP算法、Delaunay三角化法等点云数据的编辑处理,转到项目管理的具体应用场景中。同时,通过机器学习算法自动识别构件和管理信息,也成为该领域的研究热点。BIM平台的互操作问题依然突出,需要全世界的学者共同研究解决。

BIM is advanced in 3D visualization and information coordination,which is being widely adopted in the whole life cycle of building projects.Point cloud data being generated from laser scanner has characteristics in high accuracies and contactless,which is being widely implemented in many areas such as smart cities,heritages protections,and infrastructure surveys.3D laser scan-based BIM inverse modelling is a new tendency in the development of BIM technology.For objectively reflecting the current research on the BIM+3D scanning technique,this paper has used the visualization software CiteSpace to analyze 166 publication which are being included in the core collection of Web of Science in between 2011 and 2020.The results have shown that:researches are being focused on Model Reconstruction,Point Cloud Segmentation,Machine Learning,and etc.;the number of the publications are keep increasing in recent years;however,the research cooperation in this discipline is not enough;Heriot-Watt University and Poltecn Milan are the two institutions with have most publications;Bosche F.has most publication;United States and Spain are the two countries which have most publications,and then followed by China.The research focus is being transferred from point cloud editing process such as ICP algorithm and Delaunay triangulation method to specific scenario in project management.Meanwhile,research is also focusing on automatic recognition from machine learning to detect components and to manage the information.The problem of interoperability of BIM platform is still waiting to be addressed by researchers from worldwide.