Novel Approach to Inspections of As-Built Reinforcement in Incrementally Launched Bridges by Means of Computer Vision-Based Point Cloud Data

• Published in: IEEE sensors journal 2021-05, Vol.21 (10), p.11822-11833
• Main Authors: Owerko, Piotr, Owerko, Tomasz
• Format: Article
• Language: English
• Subjects: Accuracy; As-build reinforcement; Bridge construction; Bridge inspection; Bridges; Cameras; collinearity equations; Computer vision; Concrete; Concrete bridges; Data acquisition; High definition; high definition surveying; Incremental launching; incrementally launched bridges; Inspection; Laser applications; Lasers; Mathematical models; Measurement by laser beam; Photogrammetry; Post-production processing; Reinforcement; Scanners; Scanning; Sensors; Software; Three dimensional models; Three-dimensional displays; Workflow
• ISSN: 1530-437X; 1558-1748
• DOI: 10.1109/JSEN.2020.3020132

The paper presents inspection and assessment of as-built reinforcement of a selected segment of an incrementally launched (IL) concrete bridge under construction. Two novel techniques were used for data acquisition: modified photogrammetry and High Definition Surveying - a combination of terrestrial-based laser scanning, computer technology and precision control networks. The main goal of this in-situ experiment was to develop a practical, effective, yet affordable methodology for inspecting reinforcement of the above mentioned structures. In order to maintain the requirements resulting from the specifics of IL technology (seven-day cycles, 24/7 operation, two concrete pours per week), the authors have limited the maximum allowed time for data acquisition, minimized the complexity of measurement data processing and related requirements for the software and computers, and replaced the professional equipment for photogrammetry with a commonly available SLR camera. Data obtained using this method was than referred to the point cloud model obtained with a precise, state-of-the-art 3D laser scanner. The adopted mathematical model for data post-processing turned out to be effective and correct both for the analysis of data from laser scanners and photogrammetry. Presented solution adopts practical workflow with camera calibration to simplify in-situ measurement procedure achieving high accuracy standards required for reinforcement inspection. The point cloud data model obtained from photogrammetry turned out to be insufficient to assess web reinforcement thoroughly. In turn, it was possible to inspect and evaluate the reinforcement of the bottom slab, with accuracy matching the laser scanning.