Abstract
With a multitude of visual depth cues available, the visual system must often combine different cues to generate a singular impression of depth. We investigated how motion parallax (MP) and binocular disparity (BD) cues are combined to generate a single perceived depth magnitude. Our previous work (Nawrot et al., 2011, VSS) showed that depth from MP could be quantified with the transducer-corrected motion/pursuit ratio where: dMP = dθ[sup]r[/sup]/ dα[sup]e[/sup] * 0.031 f, where r= 0.417 and e = 0.193. Depth from BD (with small depth intervals) can be quantified with the distance-squared law (Cormack & Fox, 1985). The goal of this study was to determine whether a weighted linear combination of individual MP and BD cue magnitudes account for the perceived depth magnitude in the cue combination condition. Observers viewed random-dot stimuli through a mirror stereoscope that provided ocular separation for stereo stimuli and monocular viewing of the parallax stimuli. To determine the perceived depth magnitude of a translating MP stimulus, MP stimuli were compared to static stereo stimuli in a 2IFC procedure in which observers indicated greater depth magnitude. MP stimuli translated laterally generating pursuit (dα: 1.1- 3.3 d/s) while dots within the stimulus window translated laterally (peak dθ: 0.09 - 0.55 d/s). Stereo stimuli had a range of disparities (1.5 – 15 min). For each MP stimulus, a point of subjective equality (PSE) was estimated from the psychometric function, giving the amount of BD producing the equivalent magnitude of perceived depth. In the cue combination condition a fixed 6.2 min of congruent BD was added to the translating MP stimulus, and PSEs were again determined. The cue combination condition generated perceived depth magnitude greater than a linear combination of weighted MP and BD cues explains. This result suggests non-linearities in the combination of MP and BD cues (Johnson et al., 1994).
Meeting abstract presented at VSS 2013