Chapter 2
Human Eye Overview

What This Chapter is About

Simplistically, the eye does nothing more than gather visual information. However, to create images that catch the eye, understanding how the eye gathers data and the limitations of the eye can be important. To become a skillful color image specialist, you need to know how to cater to human vision complexities.

Visible Light

The previous chapter discussed color wavelengths and colors generated by additive and subtractive light. Red, for example, has a wavelength ranging from 630 to 760 nm. The human eye can sense EM energy within the 380 nm to 760 nm range, so it detects “red.” This frequency range is known as visible light. Because infrared waves have a wavelength of approximately 10^–3 to 10^–5 nm, they fall outside the range of visible light. As a result they cannot be sensed by human visual systems and require thermal imaging techniques that convert infrared energy into visible energy.

The human eye admits measured amounts of light through a small circular opening called the pupil. The pupil expands and contracts depending on the light source intensity. The light then passes through a clear lens, which, under eye muscle control, changes shape to correctly focus the light. Finally, the light strikes the retina, which converts optical power to electrical impulses that the optic nerve transmits to the brain for interpretation.

Dynamic Range

Under the right conditions, unaided eyes can see a candle burning 17 miles away. Conversely, the eye can also discern objects a million times more intense. The ability of the eye to adapt and collect information on objects with such varying levels of intensity is simply amazing.

The total range of light levels over which the eye can work is called its dynamic range. At any given instant, retinal chemical and neural mechanisms, as well as pupil size, adapt the eye to work efficiently over some small portion of the total dynamic range. Within this working range, the eye allows us to differentiate only about 100 to 200 different light intensities. The eye’s limited working range means we don’t have to have a huge intensity range at any instant (image “tones”) to exploit.

Anatomy of the Eye

It is helpful to understand what role each part of the eye plays in visual image processing. Refer to Figure 2-1 for the following discussion.

Figure 2-1  The human eye

The cornea is a dense organ that causes light to refract within the eye. One would think that this took place in the lens, but in fact, as we have said, the lens merely focuses the light at the right spot on the retina.

As we learned earlier, light refracts more for light sources with a higher frequency. When a lens refracts (bends) light to focus it, the higher frequencies (blue colors) bend more and focus in a different position than the lower frequencies (red colors). This chromatic aberration prevents a perfect focus in many instances.

To reduce chromatic aberration, a good quality camera lens contains multiple lenses, each with a different index of refraction. To provide a similar color correction, the human eye uses a lens that varies not only in shape but also in refractive index. Finally, the eye and the brain do not use color information for differentiating fine detail and this might reduce the residual effects of chromatic aberration.

The retina is a layered membrane containing neural processing cells and 130 million interspersed cells called rods and cones. The approximately 120 million rods provide humans with their night, or scotopic, vision. Rods provide achromatic perception at very low intensity light levels —which explains why moonlit scenes appear colorless. At high intensity light levels, rods excel at detecting faint contrast changes and detecting movements at the periphery (edge) of your field of view. Since image processing occurs in relatively high light levels, rods are unimportant in digital image processing.

The eight to ten million cones are spread out over the retina, with another 300,000 concentrated in one spot called the fovea. The fovea is a small, circular area that responds to our direct line of sight and is responsible for daylight, or photopic, vision. The fovea is responsible for seeing fine details and for much of our depth perception.

Most people have three types of cones. One type of cone (the L-cone) is most sensitive to long light wavelengths (yellow-red), a second type (the M-cone) to medium wavelengths (yellow-green), and the third type (the S-cone) is most sensitive to short wavelengths (blues). Together, these three cone types provide the brain color information. This explains in part why the CIE selected three particular primary colors and why most colors can be represented as a mixture of three different wavelengths at various intensities.

From the relative density of cones (color sensors) and rods, you can see that color vision is best in the fovea. The fovea also provides the best detail vision for gray scale (achromatic) intensities, because the cones also compute light intensity. As we go from the fovea towards the field of view periphery, the cone density rapidly decreases and so does visual detail. For this reason, the fovea is perceived as the “center” of our field of view and the brain directs the eye to move so visually important objects project onto the fovea.

In scotopic (night) vision, the cones are not active. This means that the best visual detail at night is achieved by looking slightly away from an object so that the object image falls not on the inactive cones but on nearby rods. In photopic (daytime) vision, the rods are still active but most of our color and gray scale visual information comes from the cones.

Having introduced the role of the eye in collecting visual information, we now move on to the most complex area of this discussion—human perception. It is the brain that interprets all the data and then reconstructs the images we see.