Cue gradient and cue density interact in the detection and recognition of objects defined by motion,contrast, or texture

The human visual system is able to extract an object from its surrounding using a number of cues. These include foreground/background gradients in disparity, motion, texture, colour, and luminance. We have investigated normal subjects' ability to detect objects defined by either motion, texture, or luminance gradients. The effects of manipulating cue density and cue foreground/background gradient on both detection and recognition accuracy were also investigated. The results demonstrate a simple additive relationship between cue density and cue gradient across forms defined by motion, luminance, and texture. The results are interpreted as evidence for the notion that form parsing is achieved via a similar algorithm across anatomically distinct processing streams.     

Perception of object unity in young infants: The roles of motion,depth, and orientation

One hundred twenty-eight 4-month-old infants were habituated to one of several displays that depicted two rod pieces above and below a box. The effects of common motion, background texture, and orientation of the rod pieces on infants' perception of unity of the partially occluded rod were examined. Infants who viewed displays in which the rod pieces were aligned and presented in front of a textured background, subsequently looked longer at a broken rod (two rod pieces separated by a gap) than at a complete rod, implying that the infants experienced the rod pieces as connected behind the box in the first display. Infants who viewed displays with no background texture, or displays in which the rod pieces were nonaligned but relatable (i.e., connected if extended behind the occluder), looked equally at the two posthabituation displays. Infants who viewed displays containing nonrelatable rod pieces looked longer at the complete rod, implying that nonrelatable edges specify disjoint objects to 4-month-olds. A threshold model, stipulating that perception of object unity is supported by multiple visual cues, is proposed to account for these results. Veridical perception of motion of display elements, depth ordering, and edge orientation are necessary, but not individually sufficient, to support young infants' perception of object unity.     

Detectability of motion as a factor in depth perception by monocular movement parallax

A display of two objects at different distances was presented to 10 observers, who were requested in two experiments to match the width of the more distant (comparison) object to the width of the nearer (standard) one under conditions permitting monocular observation and lateral head motion. The matched width of the comparison object was considered a measure of the effectiveness of movement parallax. The effectiveness of movement parallax decreases with increasing angular separation of the objects and with increasing background distance. A background without visible texture leads to a better perception of depth between two objects than a textured background The results can be explained by postulating that, whenever the detectability of motion is enhanced, i.e., the threshold for the detection of motion is lowered, the effectiveness of movement parallax as a cue to depth is increased.     

The perception of shrinking in apparent motion

Apparent motion was produced using two triangular patterns of different sizes, each exposed for 100 msec, with a 50-msec interstimulus interval and 200-msec recycle interval. The triangles were aligned either on center or on the midpoints of the bases. Experiment 1, filled, outline, and three-dot triangles were viewed over four backgrounds: a blank illuminated field, and texture gradients constructed from horizontal lines, perspective lines, or a combination of these (full texture). In Experiment 2, outline and dot triangles were presented in one of three orientations: base down, base right, and base up over a blank background. Subjects made two forced-choice responses: apparent size was categorized as shrinking or not shrinking, and apparent motion was categorized as motion in depth or motion in a fixed frontal plane. The type of alignment was the major determiner of responses. When the midpoints of the base were aligned, the predominant response described a shrinking object in a fixed-position in depth. When the centers were aligned, the predominant response described an object of constant size moving in depth.     

The perception of shrinking in apparent motion.

Apparent motion was produced using two triangular patterns of different sizes, each exposed for 100 msec, with a 50-msec interstimulus interval and 200-msec recycle interval. The triangles were aligned either on center or on the midpoints of the bases. In Experiment 1, filled, outline, and three-dot triangles were viewed over four backgrounds: a blank illuminated field, and texture gradients constructed from horizontal lines, perspective lines, or a combination of these (full texture). In Experiment 2, outline and dot triangles were presented in one of three orientations: base down, base right, and base up over a blank background. Subjects made two forced-choice responses: apparent size was categorized as shrinking or not shrinking, and apparent motion was categorized as motion in depth or motion in a fixed frontal plane. The type of alignment was the major determiner of responses. When the midpoints of the base were aligned, the predominant response described a shrinking object in a fixed position in depth. When the centers were aligned, the predominant response described an object of constant size moving in depth.     

When texture takes precedence over motion in depth perception

Both texture and motion can be strong cues to depth, and estimating slant from texture cues can be considered analogous to calculating slant from motion parallax (Malik and Rosenholtz 1994, report UCB/CSD 93/775, University of California, Berkeley, CA). A series of experiments was conducted to determine the relative weight of texture and motion cues in the perception of planar-surface slant when both texture and motion convey similar information. Stimuli were monocularly viewed images of planar surfaces slanted in depth, defined by texture and motion information that could be varied independently. Slant discrimination biases and thresholds were measured by a method of single-stimuli binary-choice procedure. When the motion and texture cues depicted surfaces of identical slants, it was found that the depth-from-motion information neither reduced slant discrimination thresholds, nor altered slant discrimination bias, compared to texture cues presented alone. When there was a difference in the slant depicted by motion and by texture, perceived slant was determined almost entirely by the texture cue. The regularity of the texture pattern did not affect this weighting. Results are discussed in terms of models of cue combination and previous results with different types of texture and motion information.     

Identifying contours from occlusion events

Surface contours specified by occlusion events that varied in density, velocity, and type of motion (rotation or translation) were examined in four experiments. As a fourth experimental factor, there were both figure-motion trials (the occluding surface moved over a stationary background) and background-motion trials (the background moved behind a stationary surface) in each experiment. Displays contained line patterns and rotary motion (Experiment 1), line patterns and translatory motion (Experiment 2), textured surfaces and rotary motion (Experiment 3), and textured surfaces and translatory motion (Experiment 4). Results indicate that contour identifications are more accurate with translation than with rotation, and that background-motion trials are generally easier than figure-motion trials. Although density in all experiments affected identifications in both background- and figure-motion trials, velocity did so in Experiment 4 only. In Experiments 1, 2, and 3, velocity affected identifications in background-motion trials but not in figure-motion trials. In Experiments 3 and 4, the rate of accretion and deletion of texture was a poor predictor of identification accuracy. These results are not consistent with previous accounts of contour perception from occlusion events, and may reflect an involvement of ocular pursuit as a mechanism for registering contour information.     

Visual mislocalisation induced by translational and radial background motion

Three experiments were conducted to explore how translational and radial background motion affected visual localisation. In experiment 1, subjects were asked to indicate the apparent position of a small spot of light flashing against a background of vertical stripes, at a varying point in time before and after rapid translational motion of the background to the left or right. When the spot was flashed before the background motion, subjects mislocalised it toward the central fixation point. An interesting finding was that this mislocalisation occurred in most cases when the background moved in the direction opposite to the visual half-field in which the spot was flashed. That is to say, a spot flashed on the right side of the fixation point was mislocalised when its background moved to the left, and not when it moved to the right; and the converse was also true. In experiment 2, concentric circles were used as the background, and moved in a contracting or expanding direction. The results indicated that mislocalisation toward the central fixation point occurred when a spot was flashed before contracting motion of the background. The same mislocalisation was observed for the spot flashed in the lower visual field, but not when it was flashed in the upper visual field (experiment 3). It is concluded that the mislocalisation is a visual illusion induced by a transient background motion toward the central fixation point.     

Effects of task-irrelevant texture motion on time-to-contact judgments

We investigated the effect of local texture motion on time-to-contact (TTC) estimation. In Experiment 1, observers estimated the TTC of a looming disk with a spiral texture pattern in a prediction-motion task. Rotation of the spiral texture in a direction causing illusory contraction resulted in a significant TTC overestimation, relative to a condition without texture rotation. This would be consistent with an intrusion of task-irrelevant local upon task-relevant global information. However, illusory expansion did not cause a relative TTC underestimation but rather also a tendency towards overestimation. In Experiment 2, a vertical cylinder moved on the frontoparallel plane. Observers judged its TTC with a finish line. The cylinder was textured with stripes oriented in parallel to its longitudinal axis. It was either not rotating, rotating such that the stripes moved towards the finish line (i.e., in the same direction as the contour), or rotating such that the stripes moved away from the finish line. Both types of texture motion caused TTC overestimation compared to the static condition. Experiment 3 showed that the different effects of task-relevant and task-irrelevant texture motion are not a mere procedural effect of the prediction-motion task. In conclusion, task-irrelevant local motion and global motion are neither averaged in a simple manner nor are they processed independently.     

Effect of visual surrounding motion on body sway in a three-dimensional environment

Unidirectional motion of a uniplanar background induces a codirectional postural sway. It has been shown recently that fixation of a stationary foreground object induces a sway response in the opposite direction (Bronstein & Buckwell, 1997) when the background moves transiently. The present study investigated factors determining this contradirectional postural response. In the experiments presented, center of foot pressure and head displacements were recorded from normal subjects. The subjects faced a visual background of 2 x 3 m, at a distance of 1.5 m, which could be moved parallel to the interaural axis. Results showed that when the visual scene consisted solely of a moving background, the conventional codirectional postural response was elicited. When subjects were asked to fixate an earth-fixed foreground (window frame) placed between them and the moving background, a consistent postural response in the opposite direction to background motion was observed. In addition, we showed that this contradirectional postural response was not transient but was sustained for the 11 sec of background motion. We investigated whether this contradirectional postural response was the consequence of the induced movement of the foreground by background motion. Although induced movement was verbally reported by subjects when viewing an earth-fixed target projected onto the moving background, the contradirectional sway did not occur. These results indicate that foreground-background separation in depth was necessary for the contradirectional postural response to occur rather than induced movement. Another experiment showed that, when the fixated foreground was attached to the head of the observer, the contradirectional sway was not observed and was therefore unrelated to vergence. Finally, results showed that the contradirectional postural response was, in the main, monocularly mediated. We conclude that the direction of the postural sway produced by a moving background in a three-dimensional environment is determined primarily by motion parallax.     

Inaccurate representation of the ground surface beyond a texture boundary

The sequential-surface-integration-process (SSIP) hypothesis was proposed to elucidate how the visual system constructs the ground-surface representation in the intermediate distance range (He et al, 2004 Perception 33 789-806). According to the hypothesis, the SSIP constructs an accurate representation of the near ground surface by using reliable near depth cues. The near ground representation then serves as a template for integrating the adjacent surface patch by using the texture gradient information as the predominant depth cue. By sequentially integrating the surface patches from near to far, the visual system obtains the global ground representation. A critical prediction of the SSIP hypothesis is that, when an abrupt texture-gradient change exists between the near and far ground surfaces, the SSIP can no longer accurately represent the far surface. Consequently, the representation of the far surface will be slanted upward toward the frontoparallel plane (owing to the intrinsic bias of the visual system), and the egocentric distance of a target on the far surface will be underestimated. Our previous findings in the real 3-D environment have shown that observers underestimated the target distance across a texture boundary. Here, we used the virtual-reality system to first test distance judgments with a distance-matching task. We created the texture boundary by having virtual grass- and cobblestone-textured patterns abutting on a flat (horizontal) ground surface in experiment 1, and by placing a brick wall to interrupt the continuous texture gradient of a flat grass surface in experiment 2. In both instances, observers underestimated the target distance across the texture boundary, compared to the homogeneous-texture ground surface (control). Second, we tested the proposal that the far surface beyond the texture boundary is perceived as slanted upward. For this, we used a virtual checkerboard-textured ground surface that was interrupted by a texture boundary. We found that not only was the target distance beyond the texture boundary underestimated relative to the homogeneous-texture condition, but the far surface beyond the texture boundary was also perceived as relatively slanted upward (experiment 3). Altogether, our results confirm the predictions of the SSIP hypothesis.     

Background stripes affect apparent speed of rotation

A gray line that rotated about its own center against a stationary background of vertical stripes appeared to double in perceptual speed as it rotated through the vertical position and thus momentarily aligned with the background. Four factors may contribute to this speed-up: (i) landmarks, in which the tip of the moving vertical line moves horizontally across the maximum number of stationary stripes; (ii) orientation repulsion of the moving line by the vertical stripes, which may distort the line's perceived position and hence its perceived speed; (iii) the orientation of an induced brightness pattern along the line; and (iv) the motion of the induced brightness pattern, which moves physically most rapidly along the line when the line is near vertical. To test these possibilities, an annulus display provided landmarks but no intersections, and this almost abolished the effect. A rotating-slit display provided an oriented, moving pattern that mimicked the induced brightness but had no landmarks, and this increased the effect. We conclude that the motion, but not the orientation, of the intersections [option (iv)] was responsible for the illusion. The fact that this motion along the length of the line affected the perceived speed of the line orthogonal to its own length indicates a failure on the part of the visual system to fully decouple tangential from radial motion.     

Kinetic occlusion: further studies of the boundary-flow cue

Previous work has demonstrated that human beings employ a processing assumption, the boundary-flow constraint, in perceiving the order of depth at an edge. Subjects perceive depth order of surfaces on the basis of the relative motions of an image boundary and a projected surface texture on either side of the boundary. In the present study, adult subjects viewed computer-generated kinematograms in which boundary-flow information provided the only cue for depth order. The results of Experiments 1 and 2 indicate that common motion between boundary and texture and differential motion between boundary and texture can independently generate the perception of ordered depths of surfaces. In Experiment 3, we examined the interaction of two processes involved in the extraction of depth order from boundary-flow displays: (1) the propagation of foreground and background surfaces from texture to boundary; and (2) the computation of depth order of surfaces on either side of the boundary. The results indicate that while the mechanism that computes depth from boundary-flow information functions reliably when the mean distance between texture and boundary is 8.1(0), surface propagation may be disrupted for distances of this magnitude.     

Seeing stereoscopic depth from disparity between kinetic edges

Traditionally, it is assumed that stereovision operates only on the positional difference (disparity) between luminance-defined features in the images in the left and the right eye. Here, I show that stereoscopic depth can be seen from disparity between edges created by relative motion of texture elements, and between edges created by correlated flicker of stationary texture elements. Luminance-based stereopsis was impossible since the texture was binocularly uncorrelated. Positional disparity of the centre of revolving patterns was not an efficient depth cue. Stereopsis from the stimuli presented here was possible even without binocular overlap of textured areas. The results provide evidence that positional disparity of kinetic edges, defined by correlated flicker or motion contrast alone, can be used as matching features to recover stereoscopic depth.     

Induced motion of a fixated target: Influence of voluntary eye deviation

Induced motion (IM) was observed in a fixated target in the direction opposite to the real motion of a moving background. Relative to a fixation target located straight ahead, IM decreased when fixation was deviated 10° in the same direction as background motion and increased when fixation was deviated 10° opposite background motion. These results are consistent with a “nystagmus-suppression” hypothesis for subjective motion of fixated targets: the magnitude of illusory motion is correlated with the amount of voluntary efference required to oppose involuntary eye movements that would occur in the absence of fixation. In addition to the form of IM studied, this explanation applies to autokinesis, apparent concomitant motion, and the oculogyral illusion. Accounts of IM that stress visual capture of vection, afferent mechanisms, egocenter deviations, or phenomenological principles, although they may explain some forms of IM, do not account for the present results.     

Interactions between spatial and spatiotemporal information in spatiotemporal boundary formation

The surface and boundaries of an object generally move in unison, so the motion of a surface could provide information about the motion of its boundaries. Here we report the results of three experiments on spatiotemporal boundary formation that indicate that information about the motion of a surface does influence the formation of its boundaries. In Experiment 1, shape identification at low texture densities was poorer for moving forms in which stationary texture was visible inside than for forms in which the stationary texture was visible only outside. In Experiment 2, the disruption found in Experiment 1 was removed by adding a second external boundary. We hypothesized that the disruption was caused by boundary assignment that perceptually grouped the moving boundary with the static texture. Experiment 3 revealed that accurate information about the motion of the surface facilitated boundary formation only when the motion was seen as coming from the surface of the moving form. Potential mechanisms for surface motion effects in dynamic boundary formation are discussed.     

Mental rotation, physical rotation, and surface media.

Subjects made mirror-normal discriminations on alphanumeric characters shown in different orientations in the picture plane. Either the characters or the background rotated during stimulus presentation in Experiments 1-3. Character rotation in the direction of mental rotation facilitated mental rotation, whereas rotation in the opposite direction inhibited it. In Experiment 4, characters were presented in different surface media so as to stimulate only one representation at a time. Mental rotation performance was similar whether the stimuli were defined by luminance, color, texture, relative motion, or binocular disparity, suggesting that mental rotation occurs at a level beyond that of the independent analyses of these different media. These results support those of Experiments 1-3 in excluding the participation of low-level motion analysis centers in the mental rotation processes.     

Perceptual depth synthesis in the visual system as revealed by selective adaptation

Selective adaptations was used to determine the degree of interactions between channels processing relative depth from stereopsis, motion parallax, and texture. Monocular adaptations with motion parallax or binocular stationary adaptation caused test surfaces, viewed either stationary binocularly or monocularly with motion parallax, to appear to slant in the opposite direction compared with the slant initially adapted to. Monocular adaptations on frontoparallel surfaces covered with a pattern of texture gradients caused a subsequently viewed test surface, viewed either monocularly with motion parallax or stationary binocularly, to appear to slant in the opposite direction as the slant indicated by the texture in the adaptation condition. No aftereffect emerged in the monocular stationary test condition. A mechanism of independent channels for relative depth perception is dismissed in favor of a view of an asymmetrical interactive processing of different information sources. The results suggest asymmetrical inhibitory interactions among habituating slant detector units receiving inputs from static disparity, dynamic disparity, and texture gradients.     

Perceived orientation in depth from line-of-sight movement

Twelve college students viewed computer-generated displays of a cross comprised of two orthogonal dotted lines, and judged the apparent in-depth orientation of the horizontal arm by positioning a horizontal bar mounted on a rotary potentiometer. The vertical arm of the simulated cross was always in the observer’s frontal plane, but the randomly textured horizontal arm was in one of nine orientations relative to the line of sight. Each observer viewed displays in which the simulated cross was, alternately, (a) stationary, (b) approaching the viewer, and (c) stationary but expanding in size. The static texture density gradient in the horizontal arm of the simulated stationary cross mediated perceived orientation in depth. Further, when motion perspective was added to the detail perspective, the impression of depth was enhanced, with the greatest enhancement obtaining at the near viewing distance. When dynamic magnification was added to the detail perspective, the impression of depth was attenuated; this effect was interpreted as an illusory case of motion perspective.     

Induced motion of a fixated target: influence of voluntary eye deviation.

Induced motion (IM) was observed in a fixated target in the direction opposite to the real motion of a moving background. Relative to a fixation target located straight ahead, IM decreased when fixation was deviated 10 degrees in the same direction as background motion and increased when fixation was deviated 10 degrees opposite background motion. These results are consistent with a "nystagmus-suppression" hypothesis for subjective motion of fixated targets: the magnitude of illusory motion is correlated with the amount of voluntary efference required to oppose involuntary eye movements that would occur in the absence of fixation. In addition to the form of IM studied, this explanation applies to autokinesis, apparent concomitant motion, and the oculogyral illusion. Accounts of IM that stress visual capture of vection, afferent mechanisms, egocenter deviations, or phenomenological principles, although they may explain some forms of IM, do not account for the present results.