Three theories for reconciling the linearity of egocentric distance perception with distortion of shape on the ground plane

The focus here is on the paradoxical finding that whereas visually perceived egocentric distance is proportional to physical distance out to at least 20 m under full-cue viewing, there are large distortions of shape within the same range, reflecting a large anisotropy of depth and frontal extents on the ground plane. Three theories of visual space perception are presented, theories that are relevant to understanding this paradoxical result. The theory by Foley, Ribeiro-Filho, and Da Silva is based on the idea that when the visual system computes the length of a visible extent, the effective visual angle is a non-linear increasing function of the actual visual angle. The theory of Durgin and Li is based on the idea that two angular measures, optical slant and angular declination, are over-perceived. The theory of Ooi and He is based on both a default perceptual representation of the ground surface in the absence of visual cues and the "sequential surface integration process" whereby an internal representation of the visible ground surface is constructed starting from beneath the observer's feet and extending outward.

Independence of egocentric and exocentric direction processing in visual space

One major issue concerning investigations of visual perception is determination of the geometrical properties of visual space. To address this issue, one must determine the relationships between geometrical features of visual space, distance, direction, angle, and size. Consistent evidence indicates that visual angle is a determinant of perceived exocentric distance. Previous evidence suggests that exocentric distance and direction are hierarchically processed, with distance preceding direction. The present study investigated the relationship between the perceptual processing of egocentric direction and exocentric direction using a task that independently provides both perceptual variables in a single judgment. The results indicated that egocentric directions were systematically overestimated, and this was not caused by either the global shape of the layout or leg length effects. Exocentric directions presented a discontinuous pattern of overestimation of smaller angles that were subtended by radial orientations and accuracy of right angles that were subtended by horizontal orientations. This could be explained by the anisotropy of visual space, a well-established phenomenon from visual space studies. The analysis of the association between the processing of these two variables revealed independence between them in which exocentric direction processing did not depend on the processing of egocentric direction processing, and vice versa. The present results and prior evidence converge on the notion of hierarchical processing in which the visual system processes the egocentric distance of objects followed by exocentric distance processing and subsequent processing. The precise positions of egocentric and exocentric directions in this chain of processing remain to be determined.