Abstract

Psychophysical evidence suggests that signals from cones are later transformed to yield an achromatic signal and two chromatically opponent ones. Although physiological observations on the lateral geniculate nucleus (LGN) of the macaque monkey have confirmed the existence of opponent transformations, these differ in important respects from the kind inferred from psychophysical evidence. In particular, the red-green chromatic pathway appears also to carry an achromatic signal, to an extent determined by the spatial properties of the stimulus. The characteristics of the yellow-blue pathway also differ from those inferred from psychophysical observations. Although signals that arise in the LGN do not meet the requirements of the psychophysical model, a linear transformation of these signals could do the required job. Analysis of the chromatic properties of neurons in striate cortex (the recipient of information from the LGN) shows that signals from the LGN are transformed substantially, so as to resemble more closely those expected on psychophysical grounds. The second-stage mechanisms inferred from psychophysical observations therefore seem to reflect the action of at least two distinct physiological transformations.

© 1986 Optical Society of America