Abstract

Such basic functions of vision as stereopsis and kineopsis can be accomplished through the use of topographic mappings of the visual field, such as those seen in the primate visual pathway. In binocular or motion parallax viewing, parallax cues to depth are contained in pairs of images that differ locally by a combination of translations, rotations, and dilations. These generalized disparities can be represented by a pair of scaler harmonic potentials defined on the visual field. These potentials are used to determine a flow of visual information that nulls these disparities so as to fuse the pair of images. The values of these potentials at each point of the visual field are computed through use of a Green’s function, which contains an explicit specification of a composition of conformal mappings of the visual field. These mappings may be interpreted as a sequence of topographic transformations of the visual field and are qualitatively similar to those seen in the lateral geniculate nucleus (LGN) and the primary and secondary cortical areas (V1 and V2) of the primate visual pathway.

© 1992 Optical Society of America