Statistical feature based target recognition

Abstract

The statistical feature-based (StaF) classifier is presented for robust high range resolution (HRR) radar moving ground target identification. The target features used for classification are the amplitude and location of HRR signature peaks. The StaF classifier was initially developed for air target identification with the primary goal of increasing classifier robustness by maintaining high performance known target identification while minimizing errors from unknown targets. Meeting this requirement is significantly more challenging than forced decision classification. Results are presented showing the performance variability of the StaF classifier with respect to feature extraction variations. More importantly, the StaF classifier performance is compared to that of the quadratic classifier. It is found that the StaF classifier performs significantly better than the quadratic at high declaration rates demonstrating that the StaF classifier can significantly reduce errors associated with unknown targets while maintaining a high probability of correct classification.