A general model for the perception of space and motion

Abstract

The perception of space and motion involves successive transformations of signals with respect to different reference systems. The visual input is coded in terms of retinal coordinates. The retinocentric values from each eye require to be unified, and to be combined with signals for eye position and movement. This egocentric reference provides a signal for the angular size, motion, or orientation of the stimulus with respect to the observer. The egocentric signals are transformed to a coordinate system that is three-dimensional-the geocentric frame of reference. Further transformations can occur at earlier levels owing to patterncentric interactions within the visual field. When the geocentric signal corresponds to the physical dimensions of space and motion, this is referred to as perceptual constancy.