Perception of relative depth interval: systematic biases in perceived depth

Area of Science:

• * Visual perception and computational neuroscience.
• * Investigating the human visual system's ability to perceive depth.

Background:

• * Binocular disparity is a key cue for depth perception.
• * Estimating depth separation typically relies on disparity and viewing distance or eye position.

Purpose of the Study:

• * To investigate the difficulty observers experience in judging depth separation across different vertical geometric configurations.
• * To explore potential explanations for these difficulties, including the empirical vertical horopter and cyclovergence.
• * To determine if metric depth is readily accessible from binocular disparity.

Main Methods:

• * Participants viewed points arranged in different vertical geometric configurations across two intervals.
• * A standard psychophysical task was used to indirectly measure cyclovergence.
• * Observed depth perception biases were compared against predictions from horopter tilt and cyclovergence.

Main Results:

• * Many observers found judging depth separation difficult with varying vertical configurations.
• * Observers who could perform the task perceived significant differences in depth between configurations.
• * Predicted biases from cyclovergence and horopter tilt did not align with observed results.

Conclusions:

• * The empirical vertical horopter and cyclovergence do not fully explain the observed depth perception biases.
• * Metric depth perception from binocular disparity is not readily available, even in simple visual scenes.
• * The human visual system may not have direct access to precise metric depth information from binocular cues alone.