Capture of stereopsis and apparent motion by illusory contours

Abstract

The removal of right-angle sectors from four circular black disks created the impression of a white "subjective" square with its four corners occluding the disks ("illusory contours"). The display in one eye was identical to that in the other except that for one eye a horizontal disparity was introduced between the edges of the cut sectors so that an illusory white square floated out in front of the paper. When a "template" ofthis stereogram was superimposed on a vertical squarewave grating (6 cycles/deg) the illusory square pulled the corresponding lines of grating forward even though the grating was at zero disparity. (Capture was not observed if the template was superimposed on horizontal gratings; the grating remained flush with the background). Analogous effects were observed in apparent motion ("motion capture"). Removal of sectors from disks in an appropriately timed sequence generated apparent motion of an illusory square. When a template of this display was superimposed on a grating, the lines appeared to move vividly with the square even though they were physically stationary. It was concluded that segmentation of the visual scene into surfaces and objects can profoundly influence the early visual processing of stereopsis and apparent motion. However, there are interesting differences between stereo capture and motion capture, which probably reflect differences in natural constraints. The implications of these findings for computational models of vision are discussed.361 Figure I shows a square defined by illusory contours (Brady

The perception of illusory contours in the hypercyclopean domain

“…Although results ostensibly similar to ours have been reported previously under dichoptic viewing conditions with a variety of stimulus patterns (e.g., Blomfield, 1973;Gregory & Harris, 1974;Harris & Gregory, 1973;Lawson, Cowan, Gibbs, & Whitmore, 1974;Ramachandran, 1986;Ramachandran & Cavanagh, 1985;Whitmore, Lawson, & Kozora, 1976), all of these studies used stimulus displays that contained monocular cues and a luminance mismatch or difference between the inducing elements and the illusory figure. In contrast, our results indicate that neither differences in luminance nor differences in temporal correlation (i.e., motion discontinuities) are necessary for the production of illusory contours.…”

On the perception of illusory contours

“…Some preliminary observations suggested that illusory contours could be enhanced considerably by superimposing them on a checkerboard pattern ( Fig. 2) so that the edges of the pacmen are lined up with the edges of the checks (Ramachandran, 1986). We found that it was easier to make perceptual judgments on those "enhanced" illusory contours than on the more usual Kanizsa-type figures, i.e.…”

“…2). We found that this produced a striking enhancement of the illusory contours and a vivid impression of a square piece of checkerboard partially occluding four black disks in the background (Ramachandran, 1986(Ramachandran, , 1992. The enhancement is seen only when the edges of the checks are collinear with the edges of the "pacmen".…”

“…Illusory contours are probably extracted fairly early in visual processing (von der Heydt, Peterhans & Baumgartner, 1985) and they are also known to powerfully constrain a number of "front-end" visual processes, such as stereo correspondence (Ramachandran, 1986) motion correspondence (Ramachandran, 1985) and shape from shading (Ramachandran, 1988). Indeed, many of these illusions (e.g.…”

Geometry and Photometry in 3D Visual Recognition

“…Second, we assume that the images are taken under similar 108 The Problem of Achieving Full Correspondence illumination conditions. The first assumption enables us to apply the geometric results described in Part I, that is, recovering the epipolar geometry between the two views.…”