Abstract
Recent studies have suggested that the visual system subtracts the optic flow pattern experienced during self-motion from the projected retinal motion of the environment to recover object motion, a phenomenon called "flow parsing". Here we investigated the extent to which the visual system accurately performs flow parsing. Two displays (83°x83°, 500ms) simulated an observer either approaching a frontal plane or translating over a ground plane providing motion parallax and static depth cues while a probe dot moved over these planes. For both displays, a vertical component (along the world Y-axis) under control of an adaptive staircase was added to the probe dot’s horizontal motion on the plane to determine when the probe motion was perceived as horizontal. The mid-point of the probe’s motion trajectory was either 2° or 4° below the focus of expansion of the flow pattern, which was fixated. If flow parsing were complete, no vertical component would be needed as truly horizontal motion on the plane would be perceived as horizontal. Instead, for all 6 participants an upward motion component was needed for the probe to perceptually move horizontally. The magnitude of this upward motion component was consistent with only 64%±4% (mean±SE) of the optic flow being subtracted at 4° eccentricity and 72%±5% at 2° eccentricity for the frontal plane condition, and 71%±3% and 81%±2% being subtracted at 4° and 2° eccentricities for the ground plane condition. For the frontal plane stimulus with a probe moving on the image plane, a condition used by Warren & Rushton (2009), only 52%±4% and 57%±4% of the optic flow was subtracted at 4° and 2° eccentricities. We conclude that while static depth cues help flow parsing for the perception of object motion during self-motion, using a purely visual strategy may not be sufficient to explain our accurate perception of object motion during self-motion.
Meeting abstract presented at VSS 2012