Color in Computer Vision: Fundamentals and Applications

Chapter 2: Color Vision

By Marcel P. Lucassen

2.1 Introduction

For any vision system, color vision is possible only when two or more light sensors sample the spectral energy distribution of the incoming light in different ways. In animal life, several instantiations of this principle are found, some of them even using parts of the electromagnetic spectrum not visible to the human eye. Human color vision is basically trichromatic, involving three types of cone photoreceptors in the retinae of our eyes. According to a number of reports, however, some women may possess tetrachromatic vision involving four photoreceptor types. Less than three functional sensors—color deficiency—is a well-known phenomenon in humans, often erroneously termed as color blindness. But apart from these two anomalies, “normal” color vision starts with the absorption of light in three cone types. Responses arising from these cones are combined in retinal ganglion cells to form three opponent channels: one achromatic (black–white) and two chromatic channels (red–green and yellow–blue). Retinal ganglion cells send off pulselike signals through the optic nerve to the visual cortex, where the perception of color eventually takes place. With the advances in neural imaging techniques, vision researchers have learned much about the specific locations of information processing in the visual cortex. How this eventually results in the perception of color and associated color phenomena in the context of other perceptual ...