The retinex theory of color, proposed by Edwin Land in the 1980s, offered an explanation of our ability to perceive color in ambient-colored environments. This effect is known as color constancy and was unexplainable by earlier theories of color vision. The term retinex is a combination of the words retina and cortex, which are the two areas responsible for the processing of visual information.
When we see an object, we automatically compare it to other objects and to our expectations (memories) of similar objects. This process is known as making inferences. For example, when you see a picture of an apple, you will make inferences to other apples you have seen, and to objects around the apple. If you were to see that same picture, but tinted a different color, you would not likely mistake the color of the apple.
For example, consider a picture of a classic fruit bowl. Apples would be red, or perhaps green. Grapes would be green and purple, bananas would be yellow, and so forth. If you immersed that picture under a bath of yellow light, you would still be able to assert to anyone that the apples were red, the bananas yellow, and so forth. However, careful measurement would show that the color of each piece of fruit would be somewhat different than in the original picture. This could be examinable by cutting out a small piece of the picture, and removing it from the context. Without context, we cannot make inferences, and color constancy disappears.
Light enters our eyes through the lens, and falls upon the retina. The retina, located in the back of your eye, contains large numbers of rods and cones. Cones are primarily responsible for color vision. There are three types of cones--one type responds best to red and yellow light, one to green and yellow, and one to blue. Different wavelengths of light activate different combinations of cones, and transmit the information through the optic nerve to the occipital lobe in the back of your brain. Note that there is no way for the retina to make inferences about colors; a given spot on your retina will respond differently to a yellow-tinted image than an nontinted one. Therefore, the retina is not directly responsible for color constancy.
Once in the occipital lobe, visual information is sent alone specific information pathways. Most of the information from cones travels along what is known as the ventral stream to the temporal lobe, which is where object identification takes place. Once here, visual information is compared to memories and context so that an object can be identified. The ventral stream is often referred to as the "What" pathway. It is here that color constancy occurs. The brain recognizes the object, and recognizes what color it should be. That information is passed along from the temporal lobe to other parts of your brain, while the specific color information that the retina originally sent is lost.
Retinex theory is based on two prominent preliminary theories of color vision. The trichromatic theory, also known as the Young-Helmholtz theory, explains the perception of light based on three types of receptors. These receptors are the red-yellow, green-yellow and blue photo-receptive cones. Thomas Young, the original founder of the theory, was intrigued by how we can perceive so many different colors. The human eye is capable of discriminating among more than 10 million colors, but we obviously can't have that many unique receptors.
According to the trichromatic theory, there are three types of receptors that experience graded responses to varying colors of light. Another important theory is the opponent-process theory, which explains effects such as negative after-images. For example, if you stare at a brightly colored picture for a minute, then stare at a blank sheet of paper, you will be able to see a negative-colored version of the picture. Similarly, if you look at a light, you will see colored spots. This is explained by the opponent-process theory, founded by Ewald Hering, which explains that color is perceived on a scale of opposites. If a particular cone is activated for long enough, it will tire out, and stop transmitting information for a short time. The brain perceives this the same way as transmitting information about the opposite color.
For example, if you stare at a red image, your red cones will tire, and stop sending the "red" signal to the brain. The lack of "red" signal is equivalent to the "green" signal, and you will see green. This is because the brain perceives differences in the amount of firing of different cones, and judges color based on those differences.
Retinex theory is a useful model, but is no longer believed to be fully accurate in describing human color vision. It is currently used in computer vision, which is the science that enables robotic equipment to determine color. Many retinex algorithms exists in industrial applications. These algorithms receive color information as input, and attempt to compare the given information to that of pure red, green and blue objects. Computer vision is being used in development of augmented reality, computer object recognition and image restoration.
- "Biological Psychology"; J.W. Kalat; 1998
- Photo Credit abstrakt farbkreis image by pmphoto from Fotolia.com
What Is Newton's Theory of Color?
Isaac Newton discovered that color is a property of light, not the other way around. He created the first "color wheel" from...
Color Theory in Kids Terms
Color theory is an exciting concept that can be sometimes overwhelming to kids. By making lessons fun when learning color theory, children...
The Trichromatic Theory of Color Vision
The trichromatic theory of color vision explains how humans are able to view different colors. Years of research have provided information on...
How to Use the GIMP Retinex Tool
GIMP has a lot of useful scripts that you can download from the GIMP registry linked in Resources below. Retinex is an...