Vasile Diaconu¹, Jocelyn Faubert²; ¹Université of Montréal, ²Université of Montréal

Introduction. Human color perception generate typical chromatic sensations from various wavelengths of the visible spectrum by exciting three broadband sensitivity photodetector types. The retinal neural network compares, differentiates and redirects signals from the three photodetectors to the cortex through several chromatic pathways. The current concept that explains the creation of the different color pathways through the retina's neural network is not compatible with its implementation in a physical instrument to characterize colors in a way similar to the visual system so the mechanism of neural color-coding of the retina is partially known. Purpose. The present study presents a physical process of de-multiplexing signals emitted by two or more types of photodetectors with a wide band of overlapping spectral sensitivities to differentiate signals corresponding to the spectral zones that the photodetectors share with each other, and signals corresponding to the distinctive spectral sensitivity zone of each photodetector. Conclusion. The model adopts two fundamental principles from retinal neuron signal processing, such as the contrast of photoreceptor signals, as well the ON and OFF properties of retina's neural network to redirect the contrasts of photoreceptor signals to different chromatic channels. The concept of this model represents a good alternative to explain the process of color coding signals from three types of retinal photodetectors to activate the different channels involved in human color perception.