Integration of ordinal and metric cues in depth processing

Area of Science:

• Visual Perception
• Computational Neuroscience
• Psychophysics

Background:

• Ordinal, configural cues (familiarity, convexity) influence perceived depth alongside binocular disparity.
• Previous research assumed luminance cues were essential for configural cues to affect figure-ground organization.
• Burge et al. (2005) used luminance edges to demonstrate configural cue influence on perceived depth.

Purpose of the Study:

• To investigate whether configural cues require luminance, contrast, or neither to influence perceived depth.
• To determine if contour saliency alone can drive the integration of ordinal and metric depth cues.

Main Methods:

• Presented random dot stereograms with varying contour types (luminance, second-order, disparity-defined) and disparities.
• Measured the influence of configural cues on perceived depth using a sufficiently large disparity pedestal (approx. 2.5 arcmin).

Main Results:

• Configural cues influenced perceived depth for second-order contours and contours defined solely by disparity.
• The effect occurred with a disparity pedestal of about 2.5 arcmin.
• Integration of ordinal and metric cues appears driven by general contour saliency, not exclusively luminance.

Conclusions:

• Configural cues can influence perceived depth even when not defined by luminance.
• Contour saliency, rather than luminance alone, plays a crucial role in integrating ordinal and metric depth information.
• Challenges the necessity of early luminance-based figure-ground organization for integrating configural and disparity cues.