Building defect inspection and data management using computer vision, augmented reality, and BIM technology

Regular inspection and management of building defects are vital for the structural integrity of buildings, but traditional manual methods often lead to inefficiencies and misjudgments in assessing defect severity. This paper presents a framework integrating computer vision, augmented reality (AR), and building information modeling (BIM) to enhance defect inspection and management. The framework includes an AR-based defect inspection (ARDI) application with interactive user interfaces and a BIM-based defect data management (DDM) platform. The ARDI application utilizes YOLOv5 + DeepSORT algorithms for effective defect detection and tracking from real-time video streams, precisely mapping defect dimensions from 2D images to 3D coordinates and synchronizing this data with BIM for uploading to the DDM platform. Two experiments confirm that the AR technology achieves centimeter-level precision, and the framework overall enhances inspection efficiency by 78.63% compared to manual methods. This advanced framework not only improves inspection efficiency but also enables managers to comprehensively manage the entire inspection process.