Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Point Cloud Analysis for Surface Defects in Civil Structures
Assessment and evaluation of damage in civil infrastructure are most often conducted visually, despite its subjectivity and qualitative nature in locating and verifying surface damage. This research presents a new workflow to analyze non-temporal point clouds to objectively locate and quantify surface damage consisting of defects, cracks, and other anomalies based solely on geometric surface feature descriptors. The detection algorithm uses three distinct surface descriptors including vertex normal, surface variation, and curvature to locate the likely damaged areas in a non-change detection approach. The quantification algorithm uses local geometric information to estimate the local in-plane and out-of-plane directions of a damaged region and ultimately the area and volume of each damaged region. The developed workflow was assessed using two synthetic datasets with known ground truths and two more real-world case studies with varying severity and mechanisms of damage. Through analyses of two synthetic datasets, it was determined that the proposed detection algorithm could accurately locate and quantify the damaged vertices, with an average error of 2%, 1%, 4%, and 18% to estimate bounding box, perimeter, areas, and volumes, respectively. Furthermore, the analysis results of two real-world datasets illustrate the developed method’s efficiency and capability to detect and quantify the potentially damaged areas in real structures. The real-world datasets also include a comparison of lidar and Structure-from-Motion (SfM) derived point clouds. Moreover, the real-world dataset analysis results emphasize the robustness of the methodology for various geometries and scales.
Civil engineering|Remote sensing
Mohammadi, Mohammad Ebrahim, "Point Cloud Analysis for Surface Defects in Civil Structures" (2019). ETD collection for University of Nebraska - Lincoln. AAI22616444.