Abstract

Neurophysiology provided color research with a wealth of single unit recordings indicating how color is coded in the early visual system [e.g., 1]. These show that when color is coded in the visual system it is encrypted together with spatial features. An obvious requirement of the coding system is that it be sufficient to contain all image information relevant to the visual system, for processing at higher levels of the visual system [2]. One of the most commonly applied hypotheses to understand the coding strategy is that the visual system is an efficient coder. This term has been used before to investigate whether the coding scheme, using building blocks recorded in physiology, is minimizing error, utilizing minimum transmission resources, combatting noise by improving signal to noise ratio and related criteria [3,4].

© 1992 Optical Society of America