BIM与图像融合的施工进度动态监测OACHSSCD
Dynamic Construction Progress Monitoring Based on BIM and Image Fusion
传统施工现场,进度管理多依赖人工,存在效率低、误差高等问题.研究提出一种融合BIM与图像识别的施工进度动态更新方法,构建涵盖"BIM解析—图像位姿重建—构件识别—体素映射—Web可视化"五阶段流程.首先,基于IFC解析BIM模型,提取构件信息;其次,采用SfM算法实现图像与BIM空间的坐标对齐;随后,构建虚拟相机,在体素化表达下完成施工状态标注;最后,基于Three.js开发Web可视化平台.在仿真场景中开展实验验证,结果表明该方法平均识别准确率达94.6％.该方法能够在虚拟场景中实现高精度、低成本的施工状态识别与可视化更新,为智慧建造背景下的施工现场动态管控提供了有效技术支撑,具有良好的应用前景.
Progress management in traditional construction sites relies on manual work,resulting in low efficiency and high error rates.A dynamic updating method for construction progress integrating BIM and image recognition was proposed,with a five-stage workflow:BIM parsing-image pose recon-struction-component recognition-voxel mapping-Web visualization.Firstly,BIM models were parsed based on IFC to extract component information.Then,the SfM algorithm was used to achieve coordi-nate alignment between image and BIM spaces.Subsequently,a virtual camera was constructed to la-bel construction states through voxel expression.Finally,a Web visualization platform was developed on Three.js.Experiments were conducted in simulated scenarios.The results show that the method a-chieves an average recognition accuracy of 94.6％.It enables high-precision and low-cost construction state recognition and visual update in virtual scenarios,provides effective technical support for dynam-ic construction site management in intelligent construction,and has good application prospects.
郑新月;车德福;徐赵东;车佳雨
东北大学 资源与土木工程学院,辽宁 沈阳 110819东北大学 资源与土木工程学院,辽宁 沈阳 110819东北大学 振动力学与智慧防灾研究院,辽宁 沈阳 110819||东南大学 中国-巴基斯坦重大基础设施智慧防灾"一带一路"联合实验室,江苏 南京 211189东北大学 振动力学与智慧防灾研究院,辽宁 沈阳 110819
建筑与水利
BIMSfM体素化建模施工进度监测进度可视化
BIMSfMvoxel-based modelingconstruction progress monitoringprogress visualiza-tion
《土木工程与管理学报》 2026 (4)
39-47,9
国家重点研发计划(2024YFC3015900)国家自然科学基金(41871310)
评论