Point Clouds of Bridge Generated by Terrestrial Laser Scanner and Images via Structure from Motion Technique: Comparison Study

• Published in: 2021 3rd International Congress on Human-Computer Interaction, Optimization and Robotic Applications (HORA) pp. 1 - 5
• Main Authors: Takhirov, Shakhzod M., Kayen, Robert E.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 11.06.2021
• Subjects: Bridges; Drone; Human computer interaction; Laser applications; Monitoring; Steel; Still Images; Structural Health Monitoring; Structure from motion; Terrestrial Laser Scanner; Three-dimensional displays
• DOI: 10.1109/HORA52670.2021.9461365

A collection of points representing an object in space is commonly called a point cloud. There are several techniques for collecting point clouds. This research is focused on a comparison study of two approaches: (1) collecting point clouds with a surveying grade terrestrial laser scanner (TLS) and (2) generating point clouds from drone-taken still images by utilizing Structure-from-Motion (SFM) technique. The paper's main objective is to compare the generated point clouds to each other and show the advantages and disadvantages of both techniques for structural health monitoring of bridges. To achieve the goal of the paper, we selected a pedestrian bridge for a comparison study. For simplicity, the study was limited to a single-span bridge. The bridge we studied is a composite steel and reinforced concrete (RC) bridge with a curved deck. A terrestrial laser scanner was used to scan the bridge from several positions, and point clouds were registered based on point-cloud to point-cloud matching. No targets were used during the collection of the point clouds by the laser scanner. In addition, still images of both bridges were taken by a drone. Based on a quantitative comparison of the results, the paper discusses the pros and cons of both approaches.