16 September 1987 Combining Synthetic Image Generation And Al To Aid Automatic Target Recognizers
Author Affiliations +
Recognition of real world targets in complex backgrounds under variable environmental conditions and operating states is presenting severe chal-lenges to designers of automatic target recognizers. This is primarily due to the lack of identifiable statistical invariants in the target/background signature. OAO Corporation has available a sophisticated signature prediction capability that can be used in conjunction with a natural language description of the recognition context to determine both the features and the feature strengths that are appropriate for the specified recognition context. Our signature prediction capability can be used to design a context adaptive target recognizer based either upon classical pattern recognition principles or more modern but less mature learning networks such as found in emerging neurocomputers.
© (1987) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
R. Keller, J. Juarez, M. Prater, T. Balaban, "Combining Synthetic Image Generation And Al To Aid Automatic Target Recognizers", Proc. SPIE 0781, Infrared Image Processing and Enhancement, (16 September 1987); doi: 10.1117/12.940537; https://doi.org/10.1117/12.940537


Thermal Modeling And IR Scene Generation
Proceedings of SPIE (September 16 1987)
Multispectral signature modeling
Proceedings of SPIE (September 16 1992)
Background clutter models for scene simulation
Proceedings of SPIE (November 15 1993)
Validation of ShipIR (v3.2): methodology and results
Proceedings of SPIE (May 04 2006)

Back to Top