VALIDATION STUDY OF A TARGET IDENTIFICATION MODEL.
Final rept. 28 Jun 61-31 Jul 65,
NORTH AMERICAN AVIATION INC COLUMBUS OH
Pagination or Media Count:
A research program was undertaken to assess the validity of a previously developed mathematical model for predicting target identification probabilities. A major validation study and three auxiliary studies were conducted. The purpose of the validation study was to compare predictions of the model with empirically obtained probabilities. Stimulus materials were generated by filming a TV monitor screen upon which was presented the scene being viewed by a TV camera as it swept over a terrain model containing various targets of differing surround and contrast. These materials were subsequently projected at three different projection speeds and viewed from each of three different distances by three groups of twenty subjects. Empirical probabilities for each of 54 flight conditions involving variations in contrast, resolution ratio, target angular subtense, and viewing time were experimentally obtained and were compared with probabilities estimated from these factors by a mathematical model of target identification performance. The three auxiliary studies were designed to investigate specific assumptions or aspects of the target identification model. Specifically, they were to determine 1 whether the predicted probability of target identification is a function of display mode static or dynamic 2 whether performance is a function of target size independent of the relative proportion of the field of view occupied by the target and 3 whether the target identification model predicts equally well for two different levels of system resolution. Author