Inspection of aging and deteriorating infrastructure such as bridges, dams, power supply infrastructures are crucial to assure their reliability as well as to estimate their remaining life. This study proposes to use small, low flying autonomous unmanned aerial vehicles, coupled with three-dimensional laser scanning, high-resolution imaging and state-of-art modeling and analysis abilities in order to provide high-precision damage detection, condition assessment and modelling of structures and infrastructure systems. In recent years, laser scanning has been shown to be effective in capturing three-dimensional geometrical information with a high degree of accuracy, which enables automatic visual damage detection and creation of high-precision information and computational models. This paper proposes a set of strategies for processing captured texture-mapped laser point cloud automatically in order to detect surface damage and large element deformation of the infrastructure system as well as to create high-fidelity computational models. Both predicted surface and determined geometric properties, which are extracted from captured texture-mapped point clouds, are used for locating, differentiating, quantifying and documenting surface damage on infrastructures. Also, high-quality volumetric finite element meshes are generated through processing the raw 3D point cloud using proposed mesh generation strategies and validated against experimental specimen. In order to show the robustness and effectiveness of these capabilities, the proposed strategies for damage detection and computational model generation are validated and tested against both synthetic and real point cloud data.
展开▼
