SAR-Based Automatic Target Recognition Avionics System

Intelligent processing of battelfield sensor data and other information provides substantial force-multiplying benefits. The timely and cost-effective leveraging of information provides a signficant advantage over adversaries not possessing comparable information processing capabilties. Among these is automatic target recognition (ATR). Under previous and current Federal funding, AbTech has developed an innovative, robust, and effective approach to automatic target recognition The approach combines the numeric processing capabilities of abductive polynomial networks with the symbolic processing power of expert systems. The result is an ATR system that intelligently processes multi-spectral imagery (i.e., laser radar, FLIR) and non-sensor information (e.g., intelligence, mission objectives) to provide a wide range of target assessment information including detection, classification, and identification. While this sytem has been developed using actual military sensor imagery, it currently operates in an engineering environment. Issues concerning its implementation on actual avionics processors have not been addressed. Therefore, ADEPT, a VHDL-based approach for determining the actual processing requirements for informatin processing algorithms, will be used to determine the implementation feasibility of the ATR system. Also, the applicability of the ASAAT methodology for synthetic aperture radar (SAR) imagery has not been considered. A complementary objective of this effort is to identify the dual-use and commercial potential for ASAAT technology.