Monitoring Urban Development Through 3D SAR Data: A Study of Southern Italy’s Structural Evolution

Urban remote sensing has become an essential tool for understanding how cities evolve over time. This thesis investigates the use of Synthetic Aperture Radar (SAR) data to monitor urban development in southern Italy, with particular attention to the detection of newly constructed buildings. The study employs long-term SAR datasets, complemented by ground truth information, to analyze structural changes in the built environment. A deep learning framework based on a res-Net architecture is implemented to identify and classify urban changes with high precision. By combining data pre-processing, noise reduction, and cross-validation with optical imagery, the methodology ensures reliable detection of building footprints. Evaluation metrics such as F1 score, IoU, and overall precision are used to assess performance. The results highlight both the potential and limitations of SAR-based change detection. Although the approach successfully identified major structural transformations, challenges remain in distinguishing subtle or small-scale modifications. However, the proposed workflow offers a scalable solution for urban monitoring, with possible applications in sustainable planning and policy making