Is depth perception of stereo plaids predicted by intersection of constraints, vector average or second-order features?

Stereo plaid stimuli were created to investigate whether depth perception is determined by an Intersection of Constraints (IOC) or Vector Average (VA) operation on the Fourier components, or by the second-order features in a pattern. Since depth discrimination of stimuli with vertical disparity is much poorer than stimuli with horizontal disparity (Westheimer, 1984), disparity may be better suited than motion for isolating the contribution of IOC, VA or second-order features. After confirming this difference between vertical and horizontal disparity using plaids, we created type II plaid stimuli where IOC predicted vertical disparity, VA predicted positive diagonal disparity and second-order features predicted negative diagonal disparity. The stimuli were comprised of a sinusoidal component with zero disparity (oriented at 0 degrees, from vertical) and a sinusoidal component with variable disparity (oriented at ±45 degrees). In a depth discrimination task, observers indicated whether they perceived the pattern as 'near' or 'far' relative to a zero disparity aperture. Observers' perception was consistent with the disparity predicted by the second-order features, indicating their dominance over IOC and VA. Additional stimuli in which these second-order features predicted vertical disparity were created to investigate whether they would dominate perception when they were no longer reliable. In this case, observers' performance was consistent with disparity predicted by an IOC or VA of the Fourier components, not the second-order features. These experiments suggest that the visual system implements more than one approach when faced with the problem of recovering depth from disparity. Although many methods may be available to the visual system, it makes use of the most reliable cue for a given condition.