Adaptive Ground Clutter Removal Algorithm for Ground Penetrating Radar Applications in Harsh Environments

In this work we present a novel adaptive ground clutter removal (AGCR) algorithm for ground penetrating radar (GPR) applications. The reduction or even removal of the disturbing ground clutter, which mostly obscures the wanted reflections of buried targets, is very important especially for the detection of shallowly buried anti-personnel mines (APMs). Most of the ground clutter removal algorithms used nowadays, show weak performances in the case of a rough ground surface or suffer from high computation complexity. This algorithm is not designed to detect flush buried APMs. Further work is needed to ensure that the algorithm can detect all APM mines from flush buried to the maximum depth of cover of 13 cm as required by the UN SOPs. The algorithm presented in this contribution is capable to remove the ground clutter resulting from undulated surfaces by estimating not only the position but also the varying signal shape. This technique enables a near optimal ground clutter reduction with a moderate computation complexity. Furthermore, the proposed AGCR algorithm preserves the characteristic impulse responses of the buried objects, which affords a following target classification based on these significant reflections. The performance improvements of the proposed AGCR algorithm compared to standard algorithms are shown on measured data. The GPR data were collected with a laboratory test set-up which is also described briefly in this work.