Speaker
Description
Identification of structural damage in civil structures is essential for ensuring safety and minimizing economic losses. Traditional methods for identifying damaged areas within the realm of structural health monitoring (SHM) often require complex numerical models and extensive sensor networks, which are costly and challenging to implement. This study introduces innovative applications of satellite remote sensing, specifically utilizing synthetic aperture radar (SAR) imagery combined with unsupervised learning, to address these challenges. This research highlights the role of distributed displacement responses from scatterer points in interferometric SAR methodologies as crucial indicators for damage identification. By analyzing all displacement responses of scatterer points. Unsupervised distance-based anomaly detection models are developed to identify damaged areas in civil structures. Scatterer points that yield anomaly scores exceeding a predefined threshold indicate potential damage. The effectiveness and practicability of this method are demonstrated through limited SAR-extracted displacement samples from a historic masonry bridge that suffered partial collapse. Results suggest that this approach is not only practical and efficient but also effective in SHM applications.
