The effect of depth information on heading judgments

Purpose. When presented with random-dot displays with little depth information, observers cannot determine their heading accuately in the presence of rotational flow without appropriate extra-retinal information (Royden, Crowell, & Banks, 1994). The addition of static depth information seems to improve heading estimation in the context of inappropriate extra-retinal signals (van den Berg & Brenner, 1994; Vishton & Cutting, 1994). The interpretation of the added-depth experiments is uncertain, however, because they used large translational flow speeds and near response probes (near probes do not distinguish between perceived linear and circular paths; Royden, 1994). We examined the effect of added-depth information on heading judgments in the presence of simulated rotations; we used translational flows and probe distances that allow a clear interpretation of the data. Methods. Stimuli simulated linear translation across a ground plane or through a 3D cloud of dots. The eyes remained stationary throughout the trial; the flow field simulated eye rotations ranging from -5 to +5 deg/sec. We compared performance with and without binocular disparity. In a second experiment, we also examined the effect of occlusion, linear perspective, and relative size by adding a regular grid of opaque walls. Results. The addition of these depth cues did not improve performance significantly. When the simulated eye rotation was greater than 1 deg/sec, observers perceived motion on a circular path in all conditions. Responses were consistent with this phenomenology. The magnitude of heading errors increased as the response probe distance increased. In addition, there was a decrease in heading errors with larger translational flow speeds. Conclusion. We found no improvement in heading estimation during simulated rotations when we added depth information to the flow displays. Some of the differences between previous results and ours can be explained by the differences in response probe distance and translational flow speed.