Recognition of High-Range-Resolution (HRR) Profile Signatures of Moving Ground Targets for Combat Identification (CID)

Limited success of current High Range Resolution (HRR) Aided Target Recognition (AiTR) approaches are largely due to the following shortcomings: 1) the heavy reliance on HRR profile features that are susceptible to measured HRR profile variability and 2) the attempt to declare targets at the highest level of specificity (e.g. T-72 instead of Tank) in all cases. The first shortcoming limits HRR AiTR performance because measured HRR profiles are susceptible to measurement noise, translational range migration (TRM), rotational range migration (RRM), speckle reflection, and self-occlusion. As a result, current approaches that use such HRR features cannot confidently discriminate between many target types. The second shortcoming aggravates the first by attempting to make highly specific target recognition declarations among ambiguous target types that typically exhibit only subtle differences in HRR profile features. In the proposed solution, Temporal Feature Correlation with Hierarchical Reasoning (TFC-HR), these shortcomings are overcome by employing: 1) HRR profile features that are robust to real-world sources of HRR profile variability and 2) hierarchical feature clustering that allows for target recognition at varying levels of specificity.