Fast Automatic Targeting Algorithms for Hyperspectral Sensors

Smart and precision-guided Army weapons systems are being developed that use long- and mid-wave-infrared (LWIR and MWIR) hyperspectral imaging (HSI) sensors for target detection and classification. New high-performance algorithms are needed to process the image signals rapidly enough for weapons targeting. A major challenge is that the IR spectral signature of a given object varies with surface temperature, orientation, range and atmospheric conditions, forming a multidimensional subspace. In Phase I Spectral Sciences, Inc. (SSI) prototyped and demonstrated novel fast and highly selective algorithms for detecting target signatures in LWIR/MWIR HSI data. SSI proposes further development, testing, automation and delivery of the algorithms in Phase II. Extensive testing will be performed with a combination of experimental and simulated data, employing state-of-the-art software tools for radiation transport (RT), atmospheric effects removal, temperature-emissivity separation, and high-radiometric-fidelity target and scene simulation. Robust subspace compression and atmospheric retrieval techniques will be used to provide fully automated target detection and classification. SSI will be assisted by Space Computer Corp. and ITT Visual Information Systems in Phase II and in Phase III commercial development of Robust Exploitation of Variable Signatures software and hardware products.