Image-Based Monitoring of Transport Infrastructures: Methods, Challenges and Applications

Visual data are reshaping structural health monitoring (SHM) for transport infrastructure. Image-based methods—from close-range photography and UAV surveys to AI-assisted vision—for inspecting bridge decks and tunnel linings, are presented as case studies from DISEG (Politecnico di Torino) and CAIT (Rutgers). The review synthesizes methodological frameworks spanning classical photogrammetry and deep learning for damage detection, with attention to multimodal fusion. The case studies apply high-resolution campaigns on deteriorated prestressed bridge decks and on tunnel segments affected by cracking and moisture ingress. We develop processing pipelines that quantify crack length/width, spall area, and color-based surface anomalies, and we correlate these indicators with independent sensing to calibrate deterioration models. We discuss inspection objectivity and repeatability, scalability to network-level surveys, and deployment constraints in difficult-access or low-visibility environments, and we outline integration pathways with digital twins to support predictive maintenance.