Subspace-Based Technique for Improvement of Shallow Buried Landmine GPR Imaging

Abstract

The clearance of anti-personnel (AP) landmines using ground penetrating radar (GPR) is challenging task because the response from shallow buried targets is usually overlapped with strong direct coupling signal and reflection from the surface. In this paper, the new approach utilizing time-domain constrained (TDC) estimator for GPR to remove undesirable clutter is presented. The proposed technique was tested with a model of PMA-1 AP landmine and compared to commonly used GPR background removal algorithms. The results are promising and reveal good spatial resolution of obtained imaging.