Suppression and Fusion in Viewing Complex Stereograms

Abstract

Julesz has demonstrated an interesting stereoscopic effect with computergenerated dot-patterns.' One member of the stereo-pair contains a random array of dots in a rectangular matrix. The second member is identical to the first except that an inner rectangular portion of the array is shifted toward one side. An inner rectangular form is not recognizable or visible within either monocular array. One does become visible when the patterns are viewed in a stereoscope. Form-perception in these patterns is based not only upon the fact that the binocularly perceived form is at a different level of depth than the background, but also occurs when the displays are rotated 90? in the stereoscope so that the dots exhibit vertical disparities. It is well known that vertical disparity does not produce a depth-effect, yet the rectangular form is seen. Also, when the inner portions of both patterns are comprised of independently selected dots, thereby eliminating any consistent transverse or vertical disparities, a form is still seen. We must, therefore, distinguish between form in depth and form without depth. Perception of binocular form without depth may be the result of either vertical or transvere disparities, or some combination of these. A possible explanation of binocular form-perception without depth is that it is due to a differential effect of binocular rivalry in the superposed patterns. There is some evidence to support this proposal. First, Wallach and Adams have shown that rivalry will occur when commensurable areas of different surface-color are superposed.2 Since dots of different surfacecolor are superposed in disparate regions of these displays, the occurrence of rivalry may be predicted. Secondly, Julesz has shown that when one member of a stereo-pair is apparently blurred, the binocularly perceived * Received for prior publication February 25, 1964, from the Sperry Rand Research Center. The author is indebted to John Duda for his assistance in the conduct of the experiments, and to Julian Hochberg and B. W. White for many helpful ideas. 1 Bela Julesz, Binocular depth perception and pattern recognition. In Colin Cherry (ed.), Information Theory, Fourth London Symposium, 1961, 212. Also see B. W. White, Stimulus-conditions affecting a recently discovered stereoscopic effect, this JOURNAL, 75, 1962, 411.