Automatic Target Recognition using High-Range Resolution Data

Abstract : A new algorithm is presented for Automatic Target Recognition (ATR) using High Range Resolution (HRR) profiles as opposed to traditional Synthetic Aperture Radar (SAR) images. ATR performance using SAR images degrades considerably in case of moving targets due to blurring caused in the cross-range domain. ATR based on HRR profiles, which are formed without Fourier transform in the cross-range, is expected to have superior performance for moving targets with the proposed method. One of the major contributions of this project so far has been the utilization of Eigen-templates as ATR features that are obtained via Singular Value Decomposition (SVD) of HRR profiles. SVD analysis of a large class of HRR data revealed that the Range-space eigenvectors corresponding to the largest singular value accounted for more than 90% of target energy. Hence, it has been proposed that the Range-space Eigen-vectors be used as templates for classification. The effectiveness of data normalization and Gaussianization of profile data in improving classification performance is also studied. With extensive simulation studies it is shown that the proposed Eigen-template based ATR approach provides consistent superior performance with recognition rate reaching 99.5% for the four class XPATCH database. This research project is being conducted in direct collaboration with the Sensors Directorate's ATR Assessment Branch, Wright Laboratories, Wright-Patt AFB, Dayton, Ohio, where it is being monitored by Dr. Rob Williams. A primary objective df this collaborative effort is to complement and augment various other ongoing research activities being conducted or supported by the Wright Labs ATR research team.