The retina as a two-dimensional detector array in the context of color vision theories and signal detection theory

A comprehensive discussion on the human retina and its color-mechanism topography as a two-dimensional photodetector array being the first phase of the human visual sensory processor is presented. The treatment is a logical deduction from optimum statistical detection theory and its application to biological sensory systems. The evolution of color vision and of the application of statistical communications theory to sensory systems and the issues common to both that merge them into a unified research thrust in vision are briefly reviewed. The basic problems of retinal topography and available data are surveyed in detail to provide the constraints for the model. The model performance is confronted with actual visual performance data, particularly threshold and discrimination data. It is concluded that the theory of statistical communications and associated techniques of signal detectability, combined with the existing empirical findings provide a most powerful means to investigate sensory communication. As argued in earlier researches of sensory communications based on the same approach, a different new concept is defined for the sensory system and the various issues raised in the problem formulation.