Purchase this article with an account.
Thomas Papathomas, Aleksandra Sherman, Anshul Jain; The role of perspective and angle polarity in perceiving 3D objects: Lessons from reverspectives. Journal of Vision 2008;8(6):454. doi: 10.1167/8.6.454.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose. Reverspectives are painted on three-dimensional surfaces with the painted perspective cues competing against the 3-D surface geometry, resulting in depth reversal and illusory motion. The objective was to examine two main factors in the illusion: perspective and angle polarity (convex/concave). Methods. We designed symmetric 3-D stimuli of two congruent trapezoids forming a convex or concave dihedral angle, to simplify the reverspective form. We used two types of stimuli: Stick figures, where perspective was signaled by the sticks' length and width; full-surface stimuli, where perspective was signaled by the trapezoids' shape and by texture gradients. We used 4 stimuli (2×2 combinations of proper/reverse perspective with convex/concave angles) and 2×2 combinations of monocular/binocular viewing with moving/stationary observers. Task was to judge the relative depth of a lateral probe with respect to a central fixation mark. The predominance of the illusory state, as inferred by the observers' response, was the measure of the illusion's strength. Results. The data pattern is very similar for the two types of stimuli. Within each type, the variations of results for the 4 stimuli are similar across all viewing conditions. An ANOVA test showed that perspective is a much stronger cue than angle polarity. Binocular viewing weakened the monocular illusion, as expected. We expected a weaker illusion for moving observers, but it was not significantly different from that of stationary observers. Conclusions. Perspective was the dominant factor in depth perception. A potential reason for observing a weak role of angle polarity is that it was confounded with and overpowered by perspective cues; we are designing a new set of stimuli to isolate the role of angle polarity. One explanation for the strong illusion for moving observers: they maintained the reverse-depth illusion, thus motion parallax signal produced illusory motion, rather than extracting veridical depth.
This PDF is available to Subscribers Only