Fast Matching for Sensor Data Exploitation Using FPGA

This proposal addresses the need to improve automatic change detection and pattern recognition in surveillance/reconnaissance systems. In those applications, large amounts of data are generated and must be stored and processed for each scene, which makesreal-time decision making a growing challenge. In the proposed Phase I project, we will study target matching algorithms which reduce the processing time. Another problem, which would be encountered in an operational imaging system, is the issue ofsignal processing hardware. The algorithms, to be studied in the proposed project, are insensitive to scale and rotation of the target in the image. The methods used in these algorithms are based on wavelet theory and genetic search paradigms. Thehardware is a compact FPGA design with externally linked memory. In the Phase II project, we will fabricate a breadboard imaging system using the algorithms and the hardware design developed and evaluated in Phase I.FPGA based integrated sensor and signalprocessing for high-speed detection and identification of targets.