Points clouds from reality capture methods, such as LIDAR, photogrammetry, and depth cameras, produce detailed geometric data of as-built structures. In raw form, however, these data are heavy and have limited use for engineers. We have developed methods for reducing point clouds in to structural “skeletons” on which Building Information Models (BIM) or Finite Element Models (FEM) can be building. Over the life of a structure, the approach can yield an evolving, as-built structural representation which could serve as a foundation for full-fledged Civil Infrastructure Digital Twins.
Computer-Aided Civil and Infrastructure Engineering, 2021.