Advanced Sensor Data Fusion

Frontier Technology, Inc. (FTI) and its research partner, University of Florida (UF), propose to develop designs for innovative discrimination algorithms for fusion of sensor (feature) and contextual information, to provide enhanced acquisition, tracking, and discrimination of threat objects in a cluttered multi-target environment. We propose to analyze the performance of the envisioned technology to support: (a) Dynamic acquisition of target state data (e.g., motion, spectral, spatial cues) from sensor output, (b) Application of multiple classifiers to target/background radar or EO/IR and target state data to identify probable target type/track/location, (c) Adaptation of classifiers to track targets given nonergodic (statistically changing) inputs, (d) Execution on small, low-power on-board processing systems Adaptive pattern selection, key to successful sensor fusion in mission- and threat-specific scenarios, will utilize FTI's TNE pattern recognition paradigm and UF's Morphological Neural Nets (MNN). Phase I will extend and analyze FTI and UF's successful, DoD-sponsored R&D for dynamic pattern recognition to develop target detection algorithms for multiple radar or EO/IR sensor data, to detect and discriminate threats from manmade or naturally-occurring clutter. Phase II will develop and test prototype image processing software to incorporate multiple sensors of differing wavebands using obscured moving and stationary targets.