07/18/2025
By Danielle Fretwell

The Francis College of Engineering, Department of Mechanical Engineering, invites you to attend a Doctoral Dissertation defense by Fabio Bottalico on: "Development of a UAV-borne stereophotogrammetry system for static and dynamic measurements of large-scale structures."

Candidate Name: Fabio Bottalico
Degree: Doctoral
Defense Date: Monday, July 28, 2025
Time: 10 a.m. to noon
Location: PER 215

Committee:

  • Advisor: Alessandro Sabato, Ph.D, Assistant Professor, Department of Mechanical and Industrial Engineering, UMass Lowell
  • Co-Advisor: Christopher Niezrecki, Ph.D, Professor, Department of Mechanical and Industrial Engineering, UMass Lowell
  • Yan Luo, Ph.D, Professor, Department of Electrical and Computer Engineering, UMass Lowell
  • Devin K. Harris, Ph.D, Department of Civil and Environmental Engineering, University of Virginia

Abstract:
Advancements in computer vision have significantly enhanced structural health monitoring (SHM) and nondestructive testing (NDT) by offering non-contact alternatives to traditional sensor-based approaches. Techniques such as Three-Dimensional Digital Image Correlation (3D-DIC) and Three-Dimensional Point Tracking (3D-PT) enable full-line-of-sight 3D measurement of geometry, displacements, and deformation using a sequence of images collected with a stereo vision system. However, their practical application to large-scale structures remains limited by several challenges. First, conventional stereo calibration requires imaging a large calibration target that occupies most of the cameras’ field-of-view (FOV), which becomes impractical as structural size increases. Second, any change in the relative position of the cameras invalidates the calibration, requiring time-consuming recalibration and constraining camera placement to rigid setups that use stiff metal bars to maintain camera relative position. This factor limits the use of stereo vision in situations where the cameras possess relative motion, such as when cameras are mounted on Unmanned Aerial Vehicles (UAVs). Third, accurate 3D-DIC and 3D-PT typically require the application of high-contrast stochastic patterns or artificial optical targets to the surface of the structure, which may be infeasible in real-world conditions due to safety, installation access, or structural constraints preventing changes to the surface structure. This dissertation presents a series of advancements aimed at mitigating these limitations in which: 1) a sensor-based stereo camera calibration method is developed that accurately measures the extrinsic parameters between the cameras in the stereo vision configuration without requiring a FOV-sized calibration target, 2) an expansion of the sensor-based calibration method is created to enable accurate 3D-DIC and 3D-PT measurements from moving cameras, including from two independent UAVs, and 3) a natural feature tracking technique is proposed and evaluated that eliminates the need for artificial surface treatment. or optical targets. Experimental campaigns in both laboratory and field environments validate the proposed methods. The research demonstrates that the sensor-based calibration achieved at least 97% agreement with traditional calibration in 3D measurements. UAV-mounted stereo systems showed 99% agreement with ground-mounted setups. Natural feature tracking achieved 97% agreement with results obtained using artificial optical targets and reference contact sensors. This research demonstrates the feasibility of using stereovision systems and UAVs to analyze large-scale structures, marking a significant step towards making computer vision techniques more user-friendly and suitable for SHM and NDT of large-scale structures.