Driving simulation provides a more ecological assessment of human heading perception in foggy conditions

Purpose. To determine the heading perception accuracy of human subjects when engaged in an active driving simulation, as compared to the primarily studied task of making conscious judgments about passively observed motion. Methods. Subjects navigated through a randomly generated forest (after Vishton & Cutting, 1995). Periodically a target appeared (see figure) and subjects steered to intercept it. Analysis of the extent to which subjects deviated from a direct path to the target provided a conservative estimate of human navigational accuracy. Subjects also passively viewed a simulated motion through the same environment and pressed a key to indicate whether their motion was aimed to the left or the right of the target object. Both tasks were performed under a variety of image blur conditions. Results. Under optimal conditions, navigational errors were slightly greater than those observed in the passive judgment task. When stimuli were blurred, however, active driving was greatly affected, while passive judgments remained relatively accurate. Conclusions. Blurring removes some class of information from an image sequence which is crucial to human visually guided navigation but which is not important to passive heading judgments. Performing these different actions may result in fundamentally distinct perceptual processing.