Target Height Estimation and Shape Reconstruction in Multiple High-Resolution Drone-Borne SAR Images Through Shadows Extraction

This paper deals with the exploitation of shadow regions in high-resolution drone-borne synthetic aperture radar (SAR) images and utilizing them for accurate target height and shape estimation. While prior research has shown the potential of SAR shadows for height estimation, these efforts have primarily focused on airborne and spaceborne platforms and lack a theoretical analysis for assessing achievable performance. In contrast, this work introduces a processing chain for target height and shape extraction and proposes a theoretical framework to evaluate system performance. The study demonstrates that the geometric flexibility and maneuverability of drones offer significant advantages in height estimation accuracy compared to conventional SAR platforms. The proposed method is validated using experimental data from a 24 GHz INRAS radar-equipped drone-borne SAR system, showing strong alignment between the extracted 2D shapes and height estimates and the theoretical predictions. Additionally, the study highlights how multi-perspective drone imagery enables the reconstruction of three-dimensional structures, further extending the potential of drone-borne SAR systems.