A human binocular-vision model based on information theory for luminance and chromaticity at isoluminance suprathreshold changes

We describe the temporal processing of the binocular-motor system at a large scale of integration. Simple visual reaction times (VRT) were measured manually for a variety of intensity-scaled stimuli for luminance and chromaticity at isoluminance changes. By comparing binocular and monocular VRT data within the context of information theory adapted to the study of neural coding, we propose a three-stage serially distributed binocular-vision model. For each type of signal, the model reveals a form of time-division multiplexing based on a monocular geometric mean as a combination rule, plus two-couple non-linear (companding) systems inserted into two communication channels that follow the Shannon–Hartley law. [Copyright &y& Elsevier]