Kinetic Depth Research Articles

Kinetic depth effect and identification of shape.

Sperling, Landy, Dosher, and Perkins (1989) argued that all previous research on the kinetic depth effect has been contaminated by stimulus artifacts, and they developed a new task that they believe can successfully overcome this problem. The present discussion offers a counter-argument that artifacts are not as pervasive in previous research as Sperling et al. imply and that their proposed method is subject to precisely the same limitations as earlier studies. Also discussed are some general guidelines of experimental design in this type of research to minimize the possibility that subjects will adopt an inappropriate response strategy that does not truly reflect their visual perceptions of three-dimensional form.

Kinetic depth effect and identification of shape.

We introduce an objective shape-identification task for measuring the kinetic depth effect (KDE). A rigidly rotating surface consisting of hills and valleys on an otherwise flat ground was defined by 300 randomly positioned dots. On each trial, 1 of 53 shapes was presented; the observer's task was to identify the shape and its overall direction of rotation. Identification accuracy was an objective measure, with a low guessing base rate of the observer's perceptual ability to extract 3D structure from 2D motion via KDE. (1) Objective accuracy data were consistent with previously obtained subjective rating judgments of depth and coherence. (2) Along with motion cues, rotating real 3D dot-defined shapes inevitably produced a cue of changing dot density. By shortening dot lifetimes to control dot density, we showed that changing density was neither necessary nor sufficient to account for accuracy; motion alone sufficed. (3) Our shape task was solvable with motion cues from the 6 most relevant locations. We extracted the dots from these locations and used them in a simplified 2D direction-labeling motion task with 6 perceptually flat flow fields. Subjects' performance in the 2D and 3D tasks was equivalent, indicating that the information processing capacity of KDE is not unique. (4) Our proposed structure-from-motion algorithm for the shape task first finds relative minima and maxima of local velocity and then assigns 3D depths proportional to velocity.

Kinetic depth effect and optic flow—I. 3D shape from Fourier motion

Fifty-three different 3D shapes were defined by sequences of 2D views (frames) of dots on a rotating 3D surface. 1. (1) Subjects' accuracy of shape identifications dropped from over 90% to less than 10% when either the polarity of the stimulus dots was alternated from light-on-gray to dark-on-gray on successive frames or when neutral gray interframe intervals were interposed. Both manipulations interfere with motion extraction by spatio-temporal (Fourier) and gradient first-order detectors. Second-order (non-Fourier) detectors that use full-wave rectification are unaffected by alternating-polarity but disrupted by interposed gray frames. 2. (2) To equate the accuracy of two-alternative forced-choice (2AFC) planar direction-of-motion discrimination in standard and polarity-alternated stimuli, standard contrast was reduced. 3D shape discrimination survived contrast reduction in standard stimuli whereas it failed completely with polarity-alternation even at full contrast. 3. (3) When individual dots were permitted to remain in the image sequence for only two frames, performance showed little loss compared to standard displays where individual dots had an expected lifetime of 20 frames, showing that 3D shape identification does not require continuity of stimulus tokens. 4. (4) Performance in all discrimination tasks is predicted (up to a monotone transformation) by considering the quality of first-order information (as given by a simple computation on Fourier power) and the number of locations at which motion information is required. Perceptual first-order analysis of optic flow is the primary substrate for structure-from-motion computations in random dot displays because only it offers sufficient quality of perceptual motion at a sufficient number of locations.

Recovery of Nonrigid Curvilinear Objects Using Optical Flow

AbstractThe kinetic depth effect which recognizes a 3‐D structure from a 2‐D information is explained from an algorithmic viewpoint among other human vision functions such as binocular vision.This paper proposes an algorithm for reconstructing the structure and motion of a 3‐D nonrigid object using the information of positions and velocities in the time sequence of images. It is assumed that the nonrigid object does not deviate significantly from rigidity. First, the object is analyzed approximately by assuming that it is rigid. Then the first‐order terms of Ct (which is an interval of two adjacent frames) are added to improve the approximation.Models of rigid and nonrigid objects were made, and information of their positions and optical flows were formed. The effectiveness of the proposed algorithm was confirmed by applying this to the data of the models.

Infants' Sensitivity to Boundary Flow Information for Depth at an Edge

We investigated infants' sensitivity to a recently discovered kinetic depth cue, boundary flow. 5-month-old infants viewed computer-generated displays in which the relation between the motion of a boundary, indicating an edge, and 2 regions of dots on either side of the boundary, indicating surfaces, provided the only information specifying the order of the 2 surfaces in depth. Infants showed a significant reaching preference for the apparently nearer region of the display. Since previous research has demonstrated that infants reach more frequently for the nearer of 2 surfaces, this result indicates that 5-month-old infants are sensitive to boundary flow information for depth at an edge.

Four-Month-Old Infants' Sensitivity to Binocular and Kinetic Information for Three-Dimensional-Object Shape

4-month-old infants were tested for sensitivity to kinetic and binocular information for 3-dimensional-object shape. The study included 2 tests: a test for sensitivity to binocular disparity and a shape perception test. The disparity sensitivity test used a preferential looking procedure developed by Held, Birch, and Gwiazda. On the basis of the results of this test, infants were assigned to disparity-sensitive and disparity-insensitive groups. In the shape perception test, a "transfer-across-depth-cues" method was employed. Infants were habituated to a rotating object whose shape was specified by kinetic information and were then presented with stationary stereograms specifying the same object and a novel-shaped object. The disparity-sensitive infants looked significantly longer at the novel object than at the familiar object, whereas the disparity-insensitive infants showed no difference in looking time to the novel and the familiar objects. The results indicate that disparity-sensitive 4-month-old infants can perceive 3-dimensional-object shape from kinetic and binocular depth information.

Parallel model of the kinetic depth effect using local computations.

This paper defines a new model for the kinetic depth effect for multidot stimuli. The calculation is performed in a cooperative-competitive network, described as a relaxation labeling process. The process involves a local iterative computation to meet best the constraints indicated by image cues to depth. Given a constraint that prefers interdot distances in three dimensions to remain constant (local rigidity), the model becomes a local parallel computation of the Ullman incremental-rigidity scheme. Several simulations of the model are described, including some in which additional cues are combined with the changing-dot-position cue.

Dynamic stereo displays for research on the recovery of three-dimensional structure

Research issues that can be addressed by combining stereoscopic and kinetic depth displays include cue conflict and recalibration, mutual constraints, specialization of cues, and differential effects of disparity and kinetic depth on other perceptions, such as size constancy. Methods of producing dynamic stereoscopic displays are reviewed, especially displays combining stereoscopic with orthographic projections of rotation in depth. A sample personal computer program in Pas-cal is provided.

Three-dimensional structure from long-range apparent motion.

Experiments are reported which show that three-dimensional structure can be perceived from two-dimensional image motions carried by objects defined solely by the differences in binocular and/or temporal correlation (ie disparity or motion discontinuities). This demonstrates that the kinetic depth effect is independent of motion detection in the luminance domain and that its relevant input comes from detectors based on some form of identity preservation of objects or features over time, ie the long-range processes of apparent motion.

Does perceptual adaptation to telestereoscopically enhanced depth depend on the recalibration of binocular disparity?

Apparent depth is significantly reduced after viewing a rotating three-dimensional form through a system of mirrors which enhances binocular disparity by optically increasing the separation between the eyes. The change in depth perception has previously been interpreted in terms of the adaptive recalibration of binocular disparity by kinetic sources of depth information. The present experiments showed that the same result is still obtained when either disparity or kinetic depth is absent during the exposure interval. The depth aftereffect failed to occur, however, when artificial pupils were used during binocular observation of the rotating form through the mirrors. These results demonstrate that the conflict between disparity and kinetic depth is neither necessary nor sufficient to produce the adaptive change in depth perception. A close association between depth and distance aftereffects was consistently found, suggesting that both types of perceptual change derive from the same underlying process. The accommodation system was implicated by the fact that the perceptual aftereffects were present following monocular exposure, but were absent under the artificial pupil exposure condition.

Perception of structure from motion: Is projective correspondence of moving elements a necessary condition?

A fundamental assumption of almost all existing computational analyses of the perception of structure from motion is that moving elements on the retina projectively correspond to identifiable moving points in three-dimensional space. The present investigation was designed to determine the psychological validity of this assumption in several different contexts. The results demonstrate that the ability of human observers to perceive structure from motion is much more general than would be reasonable to expect on the basis of existing theory. Observers can experience a compelling kinetic depth effect even when the pattern of optical motion is contaminated by large amounts of visual noise (e.g., where the signal to noise ratio is less than 0.15). Moreover, the optical deformations of shading, texture, or self-occluding contours, which would be treated as noise by existing computational models, are analyzed by human observers as perceptually salient sources of information about an object's three-dimensional form. These results suggest that the modular analyses of visual information that currently dominate the literature will have to be modified if they are to account for the high level of generality exhibited by human observers.

Perception of structure from motion: is projective correspondence of moving elements a necessary condition?

A fundamental assumption of almost all existing computational analyses of the perception of structure from motion is that moving elements on the retina projectively correspond to identifiable moving points in three-dimensional space. The present investigation was designed to determine the psychological validity of this assumption in several different contexts. The results demonstrate that the ability of human observers to perceive structure from motion is much more general than would be reasonable to expect on the basis of existing theory. Observers can experience a compelling kinetic depth effect even when the pattern of optical motion is contaminated by large amounts of visual noise (e.g., where the signal to noise ratio is less than 0.15). Moreover, the optical deformations of shading, texture, or self-occluding contours, which would be treated as noise by existing computational models, are analyzed by human observers as perceptually salient sources of information about an object's three-dimensional form. These results suggest that the modular analyses of visual information that currently dominate the literature will have to be modified if they are to account for the high level of generality exhibited by human observers.

A computational aspect of kinetic depth effect.

A method is proposed to determine the rigid structure as well as the three dimensional motion of an object from a sequence of orthographically projected images. It is assumed that the velocities as well as the positions of the points attached to the object are observable in the images. The "instantaneous rigidity condition" wich states that the relative position vector between any two points is orthogonal to the corresponding relative velocity vector is derived from the condition of rigidity. Assuming further that the rotational velocity component is constant throughout the period of observation, each of the projected relative velocity vectors is shown to move along one of the elliptical trajectories, all of which are similar to one another. The orientations of the longer axes are common to every trajectory. Then one can determine the equations of the ellipses by observing only two points in three views or three points in two views. The solution is obtained from a set of linear equations. The length of the longer axis enables one to determine the relative velocity. The instantaneous rigidity condition is then used to obtain the relative position. The special cases where the rotational axis is either perpendicular to or parallel with the image plane are discussed. The above results are discussed in relation to the relevant psychophysical observations as well as theoretical studies.

HIPS: Image processing under UNIX. Software and applications

HIPS (Human Information Processing Laboratory’s Image processing System) is a software system for image processing that runs under the UNIX operating system. HIPS is modular and flexible: it provides automatic documentation of its actions, and is relatively independent of special equipment. It has proved its usefulness in the study of the perception of American Sign Language (ASL). Here, we demonstrate some of its applications in the study of vision, and as a tool in general signal processing. Ten examples of HIPS-generated stimuli and—in some cases—analyses are provided, including the spatial filtering analysis of two types of visual illusions; the study of frequency channels with sine-wave gratings and band-limited noise; 3-dimensional perceptual reconstruction from 2-dimensional images in the kinetic depth effect; the perception of depth in random dot stereograms and cinematograms; and the perceptual segregation of objects induced by differential dot motion. Finally, examples of noise-masked, cartoon coded, and hierarchically encoded ASL images are provided.

The role of motion in infants' perception of solid shape.

Previous research has shown that infants as young as the first few months of life perceive several aspects of the three-dimensional environment. Yet we know relatively little about the visual depth information which serves as a basis for their spatial capacities. A study is reported in which a visual habituation procedure was used to examine what types of optical depth information four-month-old infants find useful in visually perceiving solid (three-dimensional) shape. Results imply that in the absence of binocular depth cues four-month-olds rely on kinetic depth information to perceive solid shape.

Luminance controls the perceived 3-D structure of dynamic 2-D displays

The accurate three-dimensional (3-D) appearance of two-dimensional (2-D) shadow projections cast by rotating 3-D objects-the kinetic depth effect (KDE)-frequently is used to visualize depth in computer-generated displays representing complex structures. According to current perceptual theories, the addition of simulated linear perspective will prevent illusory reversals of perspective associated with the KDE, due to the geometric fact that the reversed-perspective percept is inconsistent with rigid object rotation. However, the reversed, nonrigid perceptual state can be induced by correlating luminance with the represented nearness of luminous lines, thus overriding geometric cues of form and motion that are supposed to govern the perception of rigid objects in the KDE. This finding complicates theories of 3-D motion perception based on 2-D arrays, but is of practical value in the design of visual displays representing complex 3-D objects.

On the limits of perceptual complementarity in the kinetic depth effect.

In a series of experiments, we have investigated the abilities of human observers to perceive geometric properties of moving three-dimensional objects as a function of their perspective and rotational complexities. The results indicate a decreasing ability of observers to extract metric, angular, and rigid motion as the perspectives and rotations depart from parallel projections and one-parameter central rotations. In this way, quantitative limits are suggested for the principle of perceptual complementarity suggested by Shepard (1981).

On the spatio-temporal determinants of some motion effects

In a series of recent experiments we have examined the effects of varying the spatio-temporal determinants of some motion configurations on the geometric properties of the perceived motions. For both apparent motion and the kinetic depth effect (being two representative motion sources) specific spatio-temporal frequency and phase limits have been determined with respect to the various (geometrically defined) percepts. Finally, a network is proposed which enables predictions about perceived motion loci (and other geometric features), based on the equating of the observed spatio-temporal limits with the bandwidth for a filtering mechanism.

Alternative solutions to kinetic stimulus transformations.

It has been assumed that certain stimulus transformations lead directly to depth effects, that is, that such transformations are the necessary and sufficient conditions for kinetically generated depth perception. An alternative is to view such perception as the preferred solution to the problem posed by the transforming stimulus as to what even in the world is producing that transformation. In several experiments it is shown that when other solutions are supportable by the stimulus, those same transformations will no longer lead to depth perception. These other solutions become preferred on the basis of rejection of certain coincidental features of the stimulus that otherwise would have to be accepted were the kinetic depth solution to be maintained. The findings are interpreted as challenging any theory that perception is simply the direct result of stimulation or of extraction of stimulus information and as supporting the Helmkoltzian rule of perception as a construction of the most reasonable representation.

Amplitude, frequency and phase determinants of perceived rotations and rigidity in the kinetic depth effect

In a previous experiment the author has shown how perceived rotations, in the kinetic depth effect, decrease as a function of temporal frequency. It was argued that many of the ambiguous motion effects, and the temporally limited nature of the phenomenon, are due to the inability to discriminate curvature and torsion information as well as the finite time required to extract these latter sequentially dependent image parameters. In this paper we extend the investigation to consider the perception of rotations and rigidity as a function of complexity, including amplitude and phase differences between image elements. Results indicate that perceived rigidity is specifically a function of phase information, or relative motion components, and that rotations decrease as a function of complexity. In this way the curvature and torsion extraction processes are integrated with the sinusoidal nature of the image motion.