Radar Agnostic, Low Computation Synthetic Aperture Radar (SAR) Automatic Target Recognition (ATR)

During Phase I of this program, EMSI demonstrated the feasibility of substantially reducing the bias between features extracted from simulated SAR imagery of ground vehicles and features extracted from collected SAR data, and between features extracted from SAR data collected at different bands.We also demonstrated the feasibility of the bias reduction enabling high-confidence vehicle classification with a sustainable radar agnostic database.The primary objectives of the proposed Phase II program are: to extend, integrate, and fully automate the Phase I algorithms into modularized prototype software; to obtain near-real-time high-confidence radar agnostic ATR performance while significantly reducing database size and collection requirements; and to quantify both ATR performance and the level of effort needed to sustain and add to the classification database, for both Joint STARS Recap and for legacy AF platforms.