Perceived slant of binocularly viewed large-scale surfaces: a common model from explicit and implicit measures

Area of Science:

• Visual perception
• Geometrical optics
• Environmental psychology

Background:

• Perception of large-scale surface slant is known to be exaggerated.
• This slant exaggeration increases with viewing distance.

Purpose of the Study:

• To parametrically investigate the effect of viewing distance and surface slant on perceived slant.
• To quantify the relationship between actual and perceived slant in a controlled virtual environment.

Main Methods:

• Utilized a high-fidelity virtual environment to simulate large-scale surfaces.
• Employed explicit numerical estimation and implicit aspect-ratio methods to measure perceived optical slant.
• Systematically varied viewing distance and surface slant while keeping gaze direction fixed.

Main Results:

• Perceived optical slant showed a logarithmic increase with viewing distance.
• The rate of slant exaggeration was proportionally greater for shallower surfaces.
• Perceived slant was linearly related to actual slant, with a constant gain of approximately 1.5.
• The intercept of this linear relationship increased logarithmically with viewing distance.

Conclusions:

• Developed a three-parameter model that accurately fits the observed data.
• The model successfully explains previous empirical findings from real-world environments.
• Provides a quantitative framework for understanding slant perception in relation to distance and surface geometry.