Advances in imaging infrared (IIR) technology and demonstration of this technology as a capable means of target discrimination, automatic target recognition (ATR), and auto-tracking have led to the development of numerous IIR weapon systems. Although excellent analysis tools exist for describing the imaging sensors themselves, no adequate method or tools exist for characterizing the auto-detection and tracking performance capability of the sensors against targets in a variety of backgrounds. This is complicated by the fact that auto-detection and tracking techniques are difficult to characterize. It is impossible to generate a single generic metric that will accurately predict the performance of all imaging auto-trackers. Typically auto-trackers can be categorized based on their fundamental algorithm. With knowledge of the detection or tracking algorithm, an appropriate metric can be used to predict performance. This effort identifies the common detection algorithms and tracker routines and uses the fundamental algorithms as metrics. These metrics will be used to analyze real imagery from various IR sensors. A methodology for a performance metric will be developed that accurately predicts auto-detection and tracker performance and a validation plan will be developed comparing actual auto-detection and tracker systems to the metric results.
Keywords: SEEKER PERFORMANCE METRICS, INFRARED, AUTO-TRACKING, AUTONOMOUS TARGET DETECTION, CORRELATION
