RDRTEC INCORPORATED — Department of Defense SBIR Phase I: N192-059

Description

RDRTec proposes to demonstrate a testbed for the robust detection and discrimination of periscope masts using coherent SAR. This will include sensor-independent image-formation, high-false alarm rate detection capable of detecting small targets with weak signatures, followed by a set of modular screening and discrimination techniques which can be tested for their efficacy in reducing the false alarm rate and classifying remaining targets. Image formation will be performed using, Real-time Video Synthetic Aperture Radar (VSAR). Detection and discrimination will use RDRTec’s algorithms for Small Maritime Target Detection and Discrimination. Collectively the testbed is referred to as CADD (Coherent Airborne techniques for small target ¬Detection and Discrimination). It will be enhanced with new innovative techniques influenced by analytical and numerical analysis of representative, simplified simulations of radar cross sections of submarine masts as well as available data in order to develop an insight into the scattering mechanisms as a function of viewing angle, mast configuration and exposure, the surrounding sea state and radar operating frequency. Our testbed will include single and multichannel coherent processing, sparse signal separation approaches, and time-frequency analysis. Initially the testbed will be primarily a logic-based expert system, but machine learning approaches for discrimination and screening will be explored.