SAR (Synthetic Aperture Radar) target model identification method based on structure preserving dictionary learning

摘要: The invention discloses a SAR (Synthetic Aperture Radar) target model identification method based on structure preserving dictionary learning, and mainly solves the problem of poor effect when is carried out under condition low signal-to-noise ratio in prior art. comprises following implementation steps that: preprocessing, removing background clutter to obtain normalized subimage training sample image test image; carrying utilizing feature-sigh search algorithm K-SVD (K-Singular Value Decomposition) carry iterative plurality matrixes which amount equal categories; an OMP (Orthogonal Matching Pursuit) calculate spare description vector sample; calculating reconstruction error; basis minimum reconfiguration error criterion, determining result. Compared with art, disclosed by characterized that robustness improved, has ability high-accuracy ratio.