基于无人机与深度学习的道路施工进度自动化监控研究  

Automated monitoring of road construction progress based on drones and deep learning

作　　者：郭威 杨帆[2] Guo Wei;Yang Fan(China Gezhouba Group Corporation First Engineering Co.,Ltd.,Yichang 443000,China;School of Architectural Engineering,Xi’an Technological University,Xi’an 710021,China)

机构地区：[1]中国葛洲坝集团第一工程有限公司,宜昌443000 [2]西安工业大学建筑工程学院,西安710021

出　　处：《青海交通科技》2024年第4期92-97,共6页Qinghai Transportation Science and Technology

摘　　要：通过对安徽省东部某道路施工项目的案例分析,验证了无人机技术相较于传统监测方法在效率和精度上的优势。在本研究中,无人机被用于多时态影像数据的采集,再结合摄影测量技术生成高精度的三维点云模型,并采用卷积神经网络的Siamese结构实现自动化变化检测,精确计算施工进度的完成度(POC)。结果表明,该方法能够显著提高施工监测的效率和数据的可靠性,可为施工管理者提供强有力的决策支持。Through a case analysis of a road construction project in eastern Anhui Province,the advantages of drone technology in efficiency and accuracy compared to traditional monitoring methods was verified.In this study,drones were used to collect multi-temporal image data,combined with photogrammetry technology to generate a high-precision three-dimensional point cloud model,and the Siamese structure of the convolutional neural network was used to achieve automated change detection and accurately calculate the percentage of completion(POC)of construction progress.The results show that this method can significantly improve the efficiency of construction monitoring and the reliability of data,and can provide strong decision support for construction managers.

关 键 词：无人机 三维重建 施工进度监测 点云模型 卷积神经网络 变化检测 

分 类 号：TU712.3[建筑科学—建筑技术科学] TU712.1