A new SVM for distorted SAR object classification

We consider rejection and classification tests on the MSTAR (moving and stationary target acquisition and recognition) public database. We follow a benchmark procedure, which involves classification of three object classes and rejection of two confusers. This problem is difficult, since MSTAR images are specular and each target has a full 360° aspect angle range. In addition, a classifier should be able to handle object variants and depression angle differences between the training and test sets. We employ a new support vector representation and discrimination machine (SVRDM) for its excellent rejection-classification capability. A new simple registration method is used. Test results are presented and compared with those of other algorithms. The proposed method was also applied to clutter rejection and produced perfect rejection scores.

[1]  D. Casasent,et al.  Support vector machines for class representation and discrimination , 2003, Proceedings of the International Joint Conference on Neural Networks, 2003..

[2]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[3]  David Casasent,et al.  Face recognition and verification with pose and illumination variations and imposter rejection , 2005, SPIE Defense + Commercial Sensing.

[4]  Dongxin Xu,et al.  Synthetic aperture radar automatic target recognition with three strategies of learning and representation , 2000 .

[5]  Rajesh Shenoy,et al.  Feature space trajectory for distorted-object classification and pose estimation in synthetic aperture radar , 1997 .

[6]  A Mahalanobis,et al.  Intraclass and between-class training-image registration for correlation-filter synthesis. , 2000, Applied optics.

[7]  Michael Lee Bryant,et al.  SVM classifier applied to the MSTAR public data set , 1999, Defense, Security, and Sensing.

[8]  David Casasent,et al.  Classification and rejection of MSTAR data , 2004, SPIE Defense + Commercial Sensing.

[9]  Romain Murenzi,et al.  Effect of signal-to-clutter ratio on template-based ATR , 1998, Defense, Security, and Sensing.

[10]  Mohamed I. Alkanhal,et al.  Polynomial distance classifier correlation filter for pattern recognition. , 2003, Applied optics.

[11]  Qun Zhao,et al.  Support vector machines for SAR automatic target recognition , 2001 .

[12]  Michael Lee Bryant,et al.  Standard SAR ATR evaluation experiments using the MSTAR public release data set , 1998, Defense, Security, and Sensing.

[13]  Abhijit Mahalanobis,et al.  Evaluation of MACH and DCCF correlation filters for SAR ATR using the MSTAR public database , 1998, Defense, Security, and Sensing.

[14]  Michael Lee Bryant.  Target signature manifold methods applied to MSTAR dataset: preliminary results , 2001, SPIE Defense + Commercial Sensing.

[15]  Nasser M. Nasrabadi,et al.  SAR ATR using a modified learning vector quantization algorithm , 1999, Defense, Security, and Sensing.

[16]  Mark J. T. Smith,et al.  New end-to-end SAR ATR system , 1999, Defense, Security, and Sensing.

[17]  Chao Yuan,et al.  Composite filters for inverse synthetic aperture radar classification of small ships , 2002 .

[18]  Bhagavatula Vijaya Kumar,et al.  Performance of the extended maximum average correlation height (EMACH) filter and the polynomial distance classifier correlation filter (PDCCF) for multiclass SAR detection and classification , 2002, SPIE Defense + Commercial Sensing.