Academic journal article Canadian Journal of Experimental Psychology

Interactions between Symmetry and Elongation in Determining Reference Frames for Object Perception

Academic journal article Canadian Journal of Experimental Psychology

Interactions between Symmetry and Elongation in Determining Reference Frames for Object Perception

Article excerpt

Abstract Many theories of object recognition posit that objects are encoded with respect to a perceptual frame of reference. Such theories assume that factors such as symmetry and elongation are critical for the assignment of an object's primary axis, and consequently for the extraction of an object's reference frame. The present experiments directly examined the relative roles played by symmetry and elongation in the determination of an object's primary axis, and the extent to which symmetry and elongation interact with one another. We found that observers use both symmetry and elongation in extracting an object's primary axis, that the extent to which each cue dominates depends on its relative salience, and that symmetry and elongation are processed interactively, rather than in encapsulated modules.

The visual cues we receive about objects vary widely with an object's movement through space, changes in perspective or orientation, partial occlusion, or even brief disappearance. Despite wide variation in visual cues, observers often have no difficulty in correctly and rapidly recognizing objects. Most theories that account for such viewpoint invariance (e.g., Biederman, 1987; Marr 81 Nishihara, 1978) share a common assumption: Object recognition requires the derivation of a frame of reference. Objects' reference frames also can help account for the cases in which viewpoint invariance falls. If the perception of an object depends upon the frame of reference, then an object should look different depending on which frame of reference is assigned. This effect may take the form of longer recognition times for misoriented objects (e.g., Jolicoeur, 1985; Tarr & Pinker, 1989), or the even more extreme form of a complete change in an observer's percept of an object. In a classic demonstration, Mach (1886/1959) showed that the same figure could be perceived in more than one way, for example, as a square or as a diamond, depending on the orientation of the figure. Such a phenomenon makes sense within the context of visual reference frames, If, in the square/diamond demonstration, the visual system assigns a reference frame aligned with the cardinal axes in both cases, then the position of each feature with respect to the reference frame will differ depending on the object's orientation. Differences in the relative positions of features with respect to reference frames lead to different percepts of the objects - the percept of a square in one case, and of a diamond in another.

Clearly, then, reference frames seem to play a critical role in object perception; but what factors determine the derivation of an object's frame of reference? Several prominent theories of object recognition (including those of Marr & Nishihara, 1978, and Biederman, 1987) assume that both axes of elongation and of symmetry1 are important for reference frame derivation. However, relatively few studies have tested this assumption explicitly, or determined the extent to which axes derived from elongation and symmetry information might interact with one another.

One recent study that did examine the relative roles of symmetry and elongation yielded unexpected results: Observers used symmetry, but not elongation, to determine an object's primary axis (Quinlan & Humphreys, 1993). In their experiments, Quinlan and Humphreys presented participants with outline figures on paper, and asked them to physically draw in the line that "most naturally went with" each figure. Figure 1 illustrates several stimuli like those Quinlan and Humphreys used. For one class of stimuli, conjoint, each stimulus contains both an axis of symmetry and an axis of elongation, and the axes of symmetry and elongation were consistent with one another: Both were vertical or both were horizontal. In another class of stimuli, disjoint, each stimulus again contains both an axis of symmetry and an axis of elongation, but the axes of symmetry and elongation were inconsistent: When one was vertical, the other was horizontal. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.