The perceived depth from disparity as function of luminance contrast

Area of Science:

• Visual neuroscience
• Computational vision
• Perception psychology

Background:

• Stereoscopic vision systems ideally compute depth from disparity irrespective of luminance contrast.
• Existing models of cortical disparity processing do not fully account for contrast effects.

Purpose of the Study:

• To investigate how luminance contrast affects perceived depth in human stereoscopic vision.
• To determine if human depth perception exhibits contrast invariance.

Main Methods:

• Utilized random-dot stereograms to present varying levels of disparity and luminance contrast.
• Measured perceived depth by having participants match corrugation depth with a horizontal line.
• Analyzed the relationship between luminance contrast, disparity, and perceived depth.

Main Results:

• Perceived depth increased with disparity up to a critical point, then decreased.
• The maximum perceived depth and the disparity at which it occurred varied sigmoidally with luminance contrast.
• Depth perception was significantly limited at low luminance contrasts.

Conclusions:

• Human depth perception from disparity is not contrast invariant, contrary to ideal stereoscopic systems.
• Luminance contrast significantly modulates perceived depth in a lawful, non-linear manner.
• Current models of cortical disparity processing require revision to incorporate these contrast-dependent effects.