Radar detection of hidden targets

To counter long-standing reconnaissance deficiencies against target camouflage and concealment, the Radio Frequency (RF) Technology Division of Wright Laboratory has been supporting the development of Ultra-High Frequency (UHF) Synthetic-Aperture Radar (SAR) technology. The capability to detect and classify vehicles hidden under foliage and/or camouflage has been successfully demonstrated through significant airborne data collections from a 200-700 MHz, fully polarimetric SAR operated by NAWC/ERIM. Range-migration SAR processing provides sub-meter resolution imagery to support robust Automatic Target Detection and Cueing (ATD/C) processing. Lockheed Martin Tactical Defense Systems, Arizona (LMTDSAZ), is providing real-time image formation and target detection processing on commercially available parallel processors to support development of an advanced technology demonstration system for Uninhabited Aerial Vehicle (UAV) applications. An overview of the ATD/C architecture and supporting UHF SAR image data base used for algorithm development and analysis is described. Key discriminants of the detection process include efficient focus of attention screening, high discriminant dimensionality, and Bayesian Neural Network (BNN) detection. Outlined are the discriminants used, processing flow, and ongoing algorithm advances to improve detection performance and computational efficiency for typical conditions of strong tree clutter and high attenuation of concealed target signatures. Results of recent demonstrations of scaleable real-time image formation and target detection processing on Commercial Off-the-Shelf (COTS) digital processors are provided