INTEGRATED AUTONOMOUS TARGET ACQUISITION SYSTEM

Developing an integrated ATR system with a superior and significantly robust target acquisition capability requires an optimal pattern recognition methodology. HNC proposes a program to develop an optimal pattern recognition methodology based on a comparison and combination of the existing methodologies. This requires the evaluation and improvement of existing ATR algorithms and architectures. In Phase I, HNC will focus on two major tasks: comparison of different methodologies and building a prototype integrated ATR system using the optimal pattern recognition methodology. The comparison task is to identify capabilities and limitations of the existing ATR algorithms/systems. In particular, data type, requirements, assumptions, speed, reliability, and performance will be carefully reviewed. Based on this initial investigation, modifications to expand the capabilities of existing ATR algorithms/systems will be carried out. Then, an integrated ATR system with low false alarm rate and high detection probability will be built based on the existing and improved ATR algorithms/systems examined in the first task.