Abstract
One frequently reported result is that, for perceptual tasks, the amount of perceived depth is larger when motion and disparity cues are presented together than when only one of these depth cues is shown at a time. It is not known whether this increase of perceived depth in the combined condition is also found when depth estimation is aided by action. The action of reaching-to-grasp might provide additional information about the depth of target objects, thus further disambiguating the estimation of 3D depth from projected images.
We tested this through an experiment investigating the problem of integration of motion and stereoscopic depth information for two manual depth estimation tasks: one based in perception [depth estimated through fingers adjustments], and one in action [depth estimated through reach-to-grasp]. In both tasks, observers viewed a display composed of three vertical lines embedded in a cloud of random dots; two flanking lines were positioned at fixation distance with the third line midway between the two located in depth (in front of the flankers). Observers estimated five different levels of depth separation between the flankers and the central line. 3D information was provided either by binocular disparities (stereo only), image velocities (motion only) or both (stereo-motion). Haptic feedback was provided in one fourth of the trials to calibrate depth estimates and ensure that normal visual-open-loop prehension occurred even in trials lacking haptic feedback (Bingham, Coats, Mon-Williams 2007).
The results show that in the combined condition, observers perceived a larger amount of estimated depth with both the perception and action tasks.