Integration of multi-source point clouds for bridge inventory – case study

Abstract. The accurate inventory of hydro-technical infrastructure, such as bridge structures, requires the seamless integration of data capturing both above-water and submerged geometries. This study presents a comprehensive workflow for the fusion of multi-source point clouds acquired using an integrated Norbit iLidar and Winghead i77h multibeam echosounder (MBES) system. A primary focus is placed on the challenges of data acquisition in demanding environmental conditions, specifically under-bridge areas characterized by degraded GNSS reception and multipath interference. To ensure survey-grade accuracy and geometric consistency, a rigorous field calibration procedure, including a specialized LiDAR and MBES Patch Test, was implemented. The results demonstrate that precise determination of boresight angles — specifically a Roll correction of −0.3° for the LiDAR sensor—is critical to eliminating systematic offsets and "ghosting" effects in the fused model. Furthermore, the study accounts for environmental factors such as sound velocity variability (1486.5–1487.6 m/s) and beam refraction at the air-water interface. The resulting integrated point cloud, with resolutions of 1 cm (LiDAR) and 2 cm (MBES), served as the foundation for structural 3D modeling using Building Information Modeling (BIM) and Constructive Solid Geometry (CSG) approaches. The findings confirm that the proposed diagnostic and correction workflow significantly enhances the reliability of digital twins for bridge health monitoring and hydro-morphological analysis in complex engineering environments.