Real Time Signature-Based Target Detection for Future Hyperspectral Imaging Sensors

Under the proposed program we will develop a real-time processing system for hyperspectral imaging (HSI) sensors. Up to the present time the false-alarm rates for HSI sensors have been from one to three orders of magnitude too high for practical automatedwide-area HSI surveillance. The proposed work will take advantage of spectral signature information to discriminate high-value targets from natural and manmade anomalies. Specifically, we will refine the algorithms demonstrated under Phase I, program themto run in real time, and produce a deployable real-time processor which will demonstrate the effectiveness and practicality of HSI surveillance systems.