Abstract
Optic flow provides visual information about the direction of heading of a moving observer. One strategy for guiding locomotion would be to keep one's visual heading directed toward the target. Another strategy would be to aim one's physical direction of motion toward the visual target. These possibilities have been tested empirically using prism goggles or a virtual environment to dissociate visual and physical direction of motion. In previous experiments, however, conditions have forced observers to initially direct their physical heading toward the target: prior to movement onset, visual targets were displaced but no optic flow was available for comparison. The experiments reported here were designed to eliminate this confound. Subjects walked in a dynamic virtual environment to a visual target on the ground 4.5m away. Visual feedback was presented using a head mounted display system, updated in real-time based on data from an optical head tracking system. The simulated environment consisted of the target, a textured ground plane, and a cloud of random dots (to provide motion parallax). On a random subset of trials, visual space was rotated by ±15deg relative to physical space. In experiment 1, these rotations were phased in gradually over a 1m range during ongoing movements, such that no adjustment was necessary to keep visual heading directed toward the target. In experiment 2, rotations were applied prior to movement, but the appearance of the target was delayed until after subjects had already begun moving, and the target was positioned between visual and physical heading directions. In both experiments, subjects “corrected” to aim their physical heading toward the target, which required actively shifting the visual heading direction away from the target. The results suggest that visual guidance of locomotion is not based on visual heading error.
Supported by NIH grants P41-RR09283 and RO1-EY13319