基于深度学习三维重建技术的建筑施工进度管理自动化系统构建

Collaborative management of construction schedule based on deep learning 3D reconstruction technology

随着建筑工程项目管理复杂程度的不断提升,越来越多自动化、智能化的施工进度方法受到传统管理领域的关注。然而受到成本高昂且使用复杂等限制,现有的主流方法难以适用于复杂的建筑施工进度管理场景。通过对比各类三维重建技术特点,搭建了基于深度学习三维重建技术的建筑施工进度协同管理自动化系统(DLR-P),利用高速摄像头采集施工现场实时图像信息,完成由二维信息到三维信息的重建,并结合BIM动态模型技术实现对建筑施工进度的自动化管控。以重庆市巴南区某项目施工现场为例对系统进行实证研究,并对系统运行过程中的各项数据进行验证分析。结果表明:DLR-P系统平均三维重建时间为61 s,满足基本进度管理需求,能够实现建筑施工进度自动化管理,有效提升建筑施工进度管理效率。相较于目前已有的管理方式,其在运行成本及使用便捷性方面均表现出较大优势。

With the increasing complexity of construction project management,more and more automatic and intelligent construction schedule management methods are concerned by the traditional management.However,the existing mainstream methods are limited by high cost and complex use,which are difficult to apply to intricate construction schedule management scenarios.By comparing the characteristics of various kinds of 3D reconstruction technology,this study built a collaborative management system of construction schedule based on deep learning 3D Reconstruction Technology(DLR-P).By collecting the real-time image information of the construction site,the system completes the reconstruction from 2D information to 3D,and realizes the automatic control of the construction progress combined with BIM dynamic model technology.In view of the system,this study conducted a case study in the construction site of a project in Banan District of Chongqing,and analyzed the data in the process of system operation.The results show that the average 3D reconstruction time of construction schedule collaborative management system(DLR-P)based on deep learning is 61 seconds,which can meet the basic schedule management requirements,realize the automatic management of construction schedule,and effectively improve the efficiency.Compared with the existing mode,it has great advantages in the operation cost and convenience.