Singular Value Decomposition Applied to Automatic Target Recognition with High Resolution Range Profiles

Abstract

Abstract: With the development of wideband radars new applications have emerged related to this kind of sensor. That is the case of automatic target recognition based on radar imagery. In this paper a target recognition methodology based on one dimensional high resolution radar imagery is presented. 1D radar images, namely high resolution range profiles (HRRP) are comprised of range bins and contain the distribution of the scattering centers of a target providing information about target structure. In this manuscript, identification of HRRP coming from measurements of in-flight aircraft is carried out by comparison with a database of simulated HRRPs. Simulated HRRPs have a very clean signature while actual HRRPs suffer from noise and other unwanted effects making the recognition process an arduous task. In order to overcome the differences between profiles, Singular Value Decomposition (SVD) is applied to matrices of HRRP. SVD is a robust tool for the decomposition of any matrix into orthogonal basis spaces, thus, by applying SVD to the HRRP matrices and selecting the most significant singular vectors, the matrices can be split into a signal and a noise subspace. The identification algorithm proposed in this paper is based on finding the aircraft which minimizes the angle between signal subspaces. Confusion matrices for the classification of the whole test set and error rates obtained will be provided in the paper full-version. As will be shown, the use of SVD provides good recognition rates even the lack of similarity between actual and simulated profiles.