Abstract
Co-planarity in 3-D space, color and luminance contrast polarity are cues for surface integration of static images (eg He & Ooi, Perception 97; Ooi & He, ARVO 01). We now examined if luminance contrast polarity also affects surface integration of images with common motion velocity. Consider three laterally aligned similar-rectangles that are separated by horizontally moving gaps. Two possible perceptions are seen: i-the middle rectangle moves horizontally along with the two gaps; ii-the gaps are seen as occluding illusory bars moving horizontally over a single, longer rectangle, which is amodally integrated from the three lateral rectangles. Predictably, such amodal integration (due to paired L-junctions of the two rectangles) occurs only when the rectangles have similar contrast polarity. For our experiments, 2 sets of three juxtaposed rectangles, oriented 45 and 135deg, were presented for 1.5s. The middle rectangle of each set crisscrossed so that the entire stimulus appeared like an X (with moving gaps along its limbs). When the middle rectangles had similar contrast polarity with the outer rectangles, hence favoring surface integration, subjects were more likely to perceive the intersecting middle rectangles as part of a larger stationary X with occluding illusory gaps (bars) moving horizontally over it. But when the middle rectangles had opposite contrast polarity relative to the outer rectangles, they were seen as independently sliding over one another. Such effect of contrast polarity on perceived motion and surface integration decreased as the gap size increased, and when the paired L-junctions of the rectangles were eliminated by changing the flat rectangular terminals into sharp corners. Also, we showed a similar contrast polarity effect when a black diamond frame was used as an explicit occluder (creating T-junctions) in place of the gaps. Altogether our findings reinforce luminance contrast polarity as an important cue for surface integration.