Exploitation of man-made objects shadows in multi-perspective drone-borne based SAR images: preliminary experimental results

This study presents a preliminary analysis that explores the use of multi-perspective shadows in drone-borne SAR images for target characterization. Shadows in SAR images provide an advantage by offering insights into target dimensions and shapes. By leveraging system geometry, they enable the estimation of target height, adding a crucial dimension not readily apparent in conventional 2D imagery. Integrating height estimates obtained from images captured at different grazing angles can significantly enhance estimation accuracy and target characterization. This paper presents a decentralized methodology that extracts information from shadowed regions within individual images and subsequently fuses them through an averaging operation. Experimental data collected using a 24 GHz INRAS drone-borne SAR system with varying grazing angles are presented to validate the proposed approach. The results demonstrate the potential of multi-perspective shadows in reducing estimation errors and enhancing target characterization in SAR imagery.