Abstract
The unambiguous perception of depth from motion parallax (MP) relies on both retinal image motion (dθ) and extra-retinal pursuit eye movement (dα) signals, which are scaled with viewing distance (f). The motion/pursuit ratio (M/PR) describes how relative depth from motion parallax (dMP) can be quantified from these signals: dMP /f = dθ/dα. The purpose of the current study is to determine the lower limits of the M/PR parameter space in which observers can still accurately perform near/far relative depth-sign discriminations for a translating motion parallax stimulus. Stationary observers used pursuit to follow a translating random-dot MP stimulus window depicting a vertical sinusoid. Their psychophysical task was to indicate the perceived depth-phase of the sinusoid, which varied depending on the relative directions of the dθ and dα components. In separate conditions the MP stimulus window translated leftward or rightward at one of several velocities (dα: 1.3 - 25 d/s). Within conditions, peak stimulus dot velocity (dθ) varied in an interlaced staircase procedure (for different directions of stimulus window translation) down to a floor of 0.0067 d/s. The direction of stimulus dot movement, which determined stimulus depth phase, was randomized. For most observers, the minimum M/PR for the unambiguous perception of depth from motion parallax is found to be constant around 0.02, across a range of pursuit velocities. However, this changes for pursuit velocities less than ~7 d/s. Here, this minimum M/PR begins to increase as the pursuit velocity decreases. This likely reflects limits in retinal image motion processing (dθ), during pursuit eye movements. This pattern of results, especially the transition at ~ 7 d/s, is similar to previous parallactic depth thresholds documented by Ono and Ujike (2005) and Ujike and Ono (2001) when accounting for the possible differences in head-stationary and head-moving conditions.
Meeting abstract presented at VSS 2014