Dynamic Random-element Stereograms Research Articles

Interactions of spatial frequency and unequal monocular contrasts in stereopsis.

Increasing the contrast of just one eye's image degrades stereothresholds; this phenomenon is referred to as the stereo contrast paradox. In experiment one, this paradox was found to be absent in dynamic random-element stereograms; thresholds were simply limited by the lower of the two eyes' contrasts. In experiment two, in which narrowband Gabor targets were used, the paradox was found to be strongest at relatively low spatial frequencies (1 cycle deg-1). As spatial frequency was increased, the paradox gradually disappeared. At relatively high spatial frequencies (5 cycles deg-1), thresholds were generally limited by the lower of the two eyes' contrasts, as was found for the dynamic noise targets. These results demonstrate the interactions of spatial frequency and contrast in binocular image combination and yield clues as to the different roles which high and low spatial frequencies may play in stereopsis.

Direction-selective Coding of Stereoscopic (Cyclopean) Motion

This study employed a selective adaptation paradigm and investigated thresholds for direction discrimination of translational stereoscopic motion (moving binocular disparity information). The stimuli were moving arrays of randomly positioned stereoscopic discs created from disparity embedded in dynamic random-element stereograms. When discrimination thresholds were measured across a range of base directions following adaptation in a fixed direction, discrimination thresholds were maximally elevated 20–30 deg away from adaptation and reduced in the same direction as adaptation. These results are consistent with a distributed-channel model of direction coding and indicate that the direction of stereoscopic motion is encoded by adaptable direction-selective mechanisms similar to those proposed for luminance-defined motion. © 1997 Elsevier Science Ltd. All rights reserved.

Speed Discrimination of Stereoscopic (Cyclopean) Motion

This study investigated the degree to which speed of stereoscopic translational motion (i.e. moving binocular disparity information) can be discriminated in a display that minimizes position information. Observers viewed dynamic random-element stereograms depicting arrays of randomly positioned stereoscopic dots that moved bidirectionally. Two tasks were performed: a speed discrimination task and a displacement discrimination task. Across a range of conditions, speed could be discriminated under conditions in which displacement could not. Thus, speed of stereoscopic motion can be discriminated when position information is minimal. This result indicates that stereoscopic motion is sensed in a way that cannot be explained by feature tracking or by inferring the motion from memory of position and time. © 1997 Elsevier Science Ltd. All rights reserved.

Direction-selective adaptation and simultaneous contrast induced by stereoscopic (cyclopean) motion

Across four experiments, this study investigated direction-specific adaptation and simultaneous contrast induced by moving binocular disparity information (stereoscopic motion). The stimuli were moving arrays of stereoscopic dots created from dynamic random-element stereograms. Experiments 1 and 2 examined the effects of adaptation to motion in a given direction on the apparent direction of test motion. Results showed that the direction of test motion appeared repulsed away from the direction of adapting motion (repulsion aftereffect) by as much as 20 deg or more when directions of adapt and test were similar. Experiment 3 investigated transfer of the repulsion aftereffect across the stereoscopic and luminance domains by employing stereoscopic adapting motion and luminance test motion or vice versa. Results showed that the repulsion aftereffect transferred across the two stimulus domains. Experiment 4 investigated direction-specific contrast by measuring the perceived direction of two stereoscopic arrays presented simultaneously and moving in different directions. Results showed that the directions of the arrays appeared repulsed away from one another when their directions were similar. Taken together, these results suggest that the direction of stereoscopic motion is coded in the activity of directionally selective mechanisms, as is the case for luminance-domain motion. Transfer of the repulsion aftereffect between stereoscopic and luminance domains indicates the two kinds of motion perception are mediated by a common substrate.

Generation of dynamic random-element stereograms in real time with a system based on a personal computer.

A new system capable of generating static and dynamic random-element stereograms is designed and implemented. The stereogram is composed of a rectangle that can be varied in size, element density, orientation, and horizontal and vertical disparities, and can be swept at several velocities. The system is based on an inexpensive personal computer (Amiga 500), designed to work as a slave system under the control of a host computer (e.g. any personal computer). A colour monitor or two black and white monitors are suitable display devices to view the stereograms. This system is used to perform psycophysical experiments in humans and neurophysiological experiments in behaving monkeys, with the purpose of studying depth perception and binocular vision.

Disparity-tuned channels of the human visual system.

Traditionally, it has been thought that the processing of binocular disparity for the perception of stereoscopic depth is accomplished via three types of disparity-selective channels--"near," "far," and "tuned." More recent evidence challenges this notion. We have derived disparity-tuning functions psychophysically using a subthreshold summation (i.e. low-level masking) technique. We measured correlation-detection thresholds for dynamic random-element stereograms containing either one or two surfaces in depth. The resulting disparity-tuning functions show an opponent-type profile, indicating the presence of inhibition between disparity-tuned units in the visual system. Moreover, there is clear inhibition between disparities of the same sign, obviating a strict adherence to near-far opponency. These results compare favorably with tuning functions derived psychophysically using an adaptation technique, and with the tuning profiles from published single-unit recordings. Our results suggests a continuum of overlapping disparity-tuned channels, which is consistent with recent physiological evidence as well as models based on other psychophysical data.

Interocular correlation, luminance contrast and cyclopean processing

We have investigated the nature and viability of interocular correlation as a measure of signal strength in the cyclopean domain. Thresholds for the detection of interocular correlation in dynamic random element stereograms were measured as a function of luminance contrast, a more traditional measure of stimulus strength. At high contrasts, correlation thresholds were independent of contrast. At low contrasts, correlation thresholds were inversely proportional to the square of contrast. Stereothresholds were also measured as a function of both contrast and interocular correlation. At low contrasts, stereoacuity was inversely proportional to both interocular correlation and the square of contrast. These results are consistent with an inherently multiplicative mechanism of binocular combination, such as a cross-correlation of the two eye's inputs.

The cyclopean ternus display and the perception of element versus group movement

This study investigated the perception of bistable stroboscopic motion (Ternus display) with cyclopean stimuli created from retinal disparity embedded in dynamic random-element stereograms, the responses to which arise at binocular-integration levels of the visual system. To provide comparison data, observers were also tested with luminance-domain stimuli matched as closely as possible to their cyclopean counterparts. The results showed that the perception of element vs group movement was similar for both stimulus domains: element movement predominated at short interstimulus intervals (ISIs) while group movement predominated at long ISIs, and there was a tendency for a greater percentage of group movement to occur with a longer frame duration. These results cast suspicion on the interpretation of bistable motion that assumes element movement is a signature of a lower-level, short-range motion system whereas group movement is a signature of a higher-level, long-range system; both percepts are engendered at binocular-integration levels of vision.

Anisotropies in global stereoscopic orientation discrimination

Human orientation discrimination of long rectangular targets formed from dynamic random-element stereograms was assessed at four orientations (horizontal, vertical, leftand right oblique) as a function of width, disparity direction and magnitude using a temporal two-alternative, forced-choice paradigm. The results revealed the presence of an oblique effect in cyclopean orientation discrimination. In addition, observers discriminated targets with crossed disparity better than targets with uncrossed disparity, and had lower discrimination thresholds when the targets were oriented horizontally than vertically. These results demonstrate the existence of the oblique effect in the hypercyclopean domain comparable in magnitude to that present in the luminance domain.

The effect of testing method on stereoanomaly

Previous tests of stereoanomaly (a deficit in stereopsis for a given disparity direction, crossed or uncrossed) have employed stimuli in which physical contours are presented with large disparities at exposure durations too brief for eye movements. This study investigated stereoanomaly using alternative testing methods with two types of stimuli: (1) stereoscopic contours formed from dynamic random-element stereograms presented both briefly and continuously, and (2) afterimages of retinally disparate physical contours. The results from three experiments show that most subjects who are classified as stereoanomalous under conditions of brief exposure perform normally under conditions that allow long inspection periods while eliminating eye movements. These results suggest that anomalies in stereopsis previously reported may depend on the method of testing rather than on deficits in underlying neural mechanisms.

Depth separation and the Ponzo illusion.

The Ponzo illusion refers to an apparent change in length of objectively equal parallel lines induced by enclosure within an acute angle. The present study investigated this illusory change in stimulus extent as a function of the relative depth positions of the parallel lines and the inducing angle. To permit facile and unconfounded manipulation of apparent depth, the stimuli comprising the Ponzo configuration were stereoscopic contours formed from dynamic random-element stereograms. The main results were: (1) apparent depth separation exerted a strong influence on illusion magnitude; (2) this influence was asymmetrical in that illusion magnitude decreased when the inducing angle appeared in depth behind the parallel lines and increased when the inducing angle appeared in depth in front of the lines. These data are consistent with a general theory of space perception that assumes that information about depth position is processed prior to information about stimulus characteristics.

Effect of depth separation on metacontrast masking.

In most investigations of visual masking, spatially adjacent contours are located in the same depth plane, a practice derived from the assumption that depth is not a critical variable in contour formation. In nine experiments, the dependence of masking on the relative depth positions of the target and mask was examined. To permit facile manipulation of depth without introducing confounding changes in proximal stimulation, the stimuli were stereoscopic contours formed from dynamic random-element stereograms generated as anaglyphs on a projection color-television display. The target was a briefly presented Landolt C with gap position randomly varied; the mask was an annulus that surrounded the target. Masking with stereoscopic contours was robust and subject to the same spatial and temporal variables influential in masking with physical contours. Depth separation of the target and mask produced significant reductions in masking and interacted with temporal order and egocentric stimulus position. These results have implications for models of visual masking and indicate that perceived depth is important in contour formation.