Abstract
In the occlusion illusion, the visible portion of a partly occluded object appears larger than a non-occluded, physically identical region (Kanizsa, 1979). Does the visual system fill in a strip along the occluded border (the partial modal completion hypothesis), or does it perceive the occluded object as farther away and therefore larger (the apparent distance hypothesis), as in the moon illusion? Previous experiments using flat displays revealed larger illusions with stronger occlusion cues and stronger evidence for partial modal completion than apparent distance (Brooks, Lai, & Palmer, VSS-04). The present experiments used stereoscopic displays to better understand this illusion. A standard partial-circle was presented in the monitors depth plane simultaneously with a comparison partial-circle and its rectangular occluder. The depth plane of the rectangle and comparison partial-circle varied independently, each appearing in front of, co-planar with, or behind the standard. Additional control conditions contained no rectangle. Participants judged the size (Experiment 1) or depth (Experiment 2) of the comparison figure relative to the standard. Experiment 1 showed that when the comparison figure was stereoscopically in front of the adjacent rectangle, the occlusion illusion was significantly reduced, but not eliminated, probably because monocular T-junctions continue to support the perception of occlusion. Experiment 2 showed that depth judgments of the comparison figure were not measurably influenced by the rectangle, suggesting that differences in apparent distance in Experiment 1 are unlikely to have produced the occlusion illusions. The results are interpreted as consistent with the partial modal completion hypothesis. (See http://ist-socrates.berkeley.edu/~plab/projects.htm)