PSI - Issue 84

Giorgia Ghirelli et al. / Procedia Structural Integrity 84 (2026) 1047–1054

1048

1. Introduction Bridges play a fundamental role in modern transportation networks, requiring assessment of their structural integrity to ensure safe and reliable operation throughout their service life. Over the past decades, structural health monitoring (SHM) has become an essential tool for understanding the mechanical response of existing structures, especially when subjected to operational and environmental loads, ranging from system identification and model updating (Ranieri et al. (2020); Ponsi et al. (2022)) to health assessment and damage detection (Comanducci et al. (2016); Ponsi et al. (2023)). Traditional dynamic monitoring systems rely primarily on accelerometers, which can provide accurate high frequency measurements and have long been considered the standard solution for vibration-based assessments. Nevertheless, these sensors also present several limitations. The installation of accelerometers often requires access to difficult locations, the deployment of cables or wireless networks, and coordination with infrastructure managers to ensure safe working conditions. Furthermore, when displacements need to be derived from acceleration data, the process involves a numerical double integration. This step can accumulate errors and amplify noise, which may compromise the reliability of the overall procedure. In recent years, contact-less technology for displacement monitoring represents a promising alternative to traditional sensor networks. In civil monitoring applications emerge the use of global navigation satellite systems (GNSS) (Poluzzi et al. (2019)), satellite remote sensing (Bassoli et al. (2023)), terrestrial radar interferometry (Castagnetti et al. (2019); Guerzoni et al. (2023)), and vision-based techniques (Fradelos et al. (2020)). Among them, the latter leverages digital imaging devices to measure the displacement of selected points of a structure directly from video recordings, reducing installation time and minimizing interference with the normal operation of the monitored structure. Consumer-grade cameras, now capable of delivering high-resolution and high-frame-rate footage (Xu et al. (2019)), can be deployed rapidly at limited cost, making vision-based approaches particularly attractive for temporary or periodic SHM campaigns. The ability to simultaneously monitor multiple points within the field of view represents an additional advantage, allowing for a richer spatial characterization of structural behavior without the need for extensive cabling or sensor arrays. In addition, only distinctive targets placed at selected locations are required, either in the form of artificial target or structural details that naturally provide identifiable features, such as sharp corners or bolts (Dong et al. (2020)). As a result, physical installations are significantly reduced or even avoided, an aspect that is particularly advantageous when the structure is difficult or unsafe to access or when it belongs to the cultural heritage. Despite these advantages, vision-based monitoring remains sensitive to environmental and operational uncertainties (Ye et al. (2016)). Factors such as camera stability, lighting variability, and perspective distortions may significantly affect measurement accuracy, thereby requiring careful methodological design to ensure reliable displacement estimation. Moreover, vision-based monitoring approaches have been mainly investigated through theoretical analyses and controlled laboratory experiments (Zona (2020)). The few and latest applications to real cases are comprehensively reviewed in the works of Spencer et al. (2019), Dong and Catbas (2020), and Zona (2020), comprising tests on bridges (Feng and Feng (2017); Chen et al. (2018)) and footbridges (Xu et al. (2018); Lydon et al. (2019)). Finally, the use of artificial targets, such as high-contrast checkerboard-pattern markers, can significantly improve tracking reliability, but proper detection procedures remain essential to ensure sub-pixel accuracy. In this study, a vision-based monitoring strategy is applied to a reinforced concrete railway arch bridge. High resolution commercial cameras were mounted beneath a selected span to record the dynamic response induced by train passages. Checkerboard targets are installed at key locations on the bridge to measure both the horizontal and vertical displacements needed to characterize its behaviour, ensuring reliable detection and sub-pixel tracking. Two different algorithms are compared for evaluating the motion of the targets, providing insights into the accuracy and robustness of the vision-based measurements. Moreover, finite element analyses are performed to investigate the structural behavior under train loads. The combination of experimental measurements and numerical simulations provides a comprehensive framework for assessing the dynamic performance of the bridge, enabling a more informed understanding of the interaction between the structure and operational loads.