Automatic Target Cueing and Operator Performance with Enhanced APG-70 Synthetic Aperture Radar Imagery

摘要

The present study examined automatic target cueing (ATC) and target localization performance using 4 ft resolution synthetic aperture radar (SAR) imagery collected during the Theater Missile Defense Eagle Smart Sensor ATC (TESSA) program at Eglin Air Force Base, FL. Twelve observers viewed 360 SAR images that varied in terms of aiding (unaided versus aided) and the amount and type of background foliage ('open,' 'treeline,' and 'sparse' sites). All images contained the target to be detected, a transporter-erector-launcher (TEL), as well as two support vehicles. In the aided condition, each image was presented with four cue boxes to identify the ATC's target designations. The same set of 180 images was also presented in a separate unaided block without cues. After viewing each SAR patch map, participants identified the location of the TEL and rated their confidence in their decision. Overall, ATC cueing at the 4 ft resolution enhanced operators' confidence in their decision making but did not alter their localization accuracy, perceptual sensitivity (d'), or speed relative to the unaided condition. Further analysis revealed, however, that a critical determinant of operator performance and confidence was the 'reliability' of the ATC. If all of the ATC's cues were false alarms, performance was worse than if no aiding had been provided at all. On the other hand, performance was most effective when the majority of the cues were centered over man-made vehicles. Finally, clutter in the form of background foliage degraded localization accuracy, sensitivity, reaction time, and operator confidence. As revealed by the present findings, when unaided performance is exacerbated by factors such as high background clutter, the assistance of a reliable ATC may be most beneficial.