Monocular Vision-Based Target-free Three-dimensional (3D) Vibration Displacement Measurement for Civil Structures

Abstract

This dissertation introduces novel monocular vision-based approaches for 3D displacement measurement in civil engineering structures. They enable accurate measurement of 3D displacements using just a static monocular camera. The cornerstone of this innovation lies in the utilization of advanced deep learning neural networks, which have the capability to extract 3D information from a single input image.