Automatic Feature Combined Track-Detect-Localize Technique

A progressive system development is proposed which uses Bayesian methodology for combined detect, track, localize, and classify. The core paradigm focuses on detection followed by various estimation schemes to extract information on contact type, track, and location from the detector. A rapid prototype development using the Generic Multipurpose Processor (GMP) will provide optimal detection of a dynamic target. Further enhancements, including classification of contacts as either surface or sub-surface, and contact identification using a database of signatures will be investigated to determine feasibility.