2026 Volume 7 Issue 1 Pages 76-86
The aging of public infrastructure has emerged as a critical challenge across Japan, with the structural deterioration of bridges being a primary concern. To address this, this study focuses on assessing structural health by measuring mid-span deflection. While image processing technologies have been previously applied to this task, this paper proposes a simplified measurement technique that requires only the placement of a monocular camera on the bridge deck. The method calculates deflection based on the displacement of captured scenery. This paper outlines the fundamental principles of the measurement system and evaluates its feasibility through experiments under controlled, ideal conditions. Furthermore, the effectiveness of the method is validated through field measurements on an actual pedestrian bridge. Finally, the study explores the integration of deep learning-based object detection for the automated identification of target features within the captured images.
