Acquisition of Sea Targets. Part 1: Observer Performance and 'ACQUIRE' ModelPredictions for Air-to-Surface FLIR Imagery

摘要

In this study, two laboratory experiments were carried out to test how well humanobservers using an electrooptical viewing device, are able to identify or classify sea targets. Such knowledge is of interest to evaluate the applicability of target acquisition models for a sea environment. Most models are designed and tested for ground targets and backgrounds, and their reliability for acquisition of sea targets is unknown. In the first experiment, which is reported here, observer performance was measured on real thermal infrared (FLIR) imagery that was collected on ORION flights. Experienced observers from several Royal Dutch Navy ORION and LYNX helicopter squadrons participated in this experiment. First, some rules of thumb were deduced from the data. For example, with the sensor that was used in the experiment and good atmospheric conditions, an S-frigate in side view may be classified (50% correct) at 14 km and identified at 7 km. For the same target in front view these ranges are about 5 km and 2 km, respectively. For the Tydeman and a Fishing Boat in side view, the 50% correct classification and identification ranges are about 7 and 4 km. Second, the results of the experiment were used to test a widely used TA model, ACQUIRE. Important differences were found between measured and predicted performance. On average, the most recent version of ACQUIRE overestimates acquisition ranges by a factor of 2.