A Multi-Processor Embedded Data Processor for Multi-Object Detection & Tracking

The Aeronix Phase I task will demonstrate the application of an innovative parallel processing computer architecture and parallel processing programing envirorment to a class of problems associated with the real time detection, recognition, acquisition and tracking of multLple objects in a high rate image sensor environment. The Phase I task will examine the problems associated with the multiplicity of objects in a large image data space and present a ccmprehensive process for the collection filtering, transformation and analysis of these objects across multiple processing elements. The Phase I task will map the Kinetic Energy Anti-Satellite program requirements on the Aeronix proprietary Embedded Data Processor (EDP) Architecture. The EDP is a space based Multi Instruction Multi Data Stream (MIND) parallel processing architecture for embedded data processing applications. The EDP offers significant increases in performance (up to 1000's of HIPS) over conventional architectures while improving overall system reliability and provLding inherently high levels of fault tolerance.