Non-Imaging Infrared Spectral Target Detection

Automatic detection of time-critical mobile targets using spectral-only infrared radiance data is explored. A quantification of the probability of detection, false alarm rate, and total error rate associated with this detection process is provided. A set of classification features is developed for the spectral data, and these features are utilized in a Bayesian classifier singly and in combination to provide target detection. The results of this processing are presented and sensitivity of the class separability to target set, target configuration, diurnal variations, mean contrast, and ambient temperature estimation errors is explored. This work introduces the concept of atmospheric normalization of classification features, in which feature values are normalized using an estimate of the ambient temperature surrounding the target being observed and applying the Plauck radiation law with those estimates. This technique is demonstrated to reduce the total error rate associated with classification processing to less than one-fourth of that observed using non-normalized features. Classification testing of spectral field measurements made on an array of U.S. and foreign military assets reveal a total error rate near 5% with a 95% probability of detection and a concurrent false alarm rate of 4% when a single classification feature is employed. Multiple feature classification on the same data yields detection probabilities near 97% with a concurrent false alarm rate of 2.5%. Sensitivity analysis indicates that the probability of detection is reduced to 70-75% in the hours preceding daylight, and that for the total error rate to be less than 10%, the target-to-background mean contrast must be greater than 0.1. Analysis of the atmospheric normalization technique reveals that in order to keep the total error rate less than 10%, the ambient temperature must be estimated wit