An image gradient is a term used to describe a continuous shift from light to dark, or dark to light, in an image. This shift refers specifically to the intensity of a single color, meaning that a single color will go from white, maximum intensity, to black, minimum intensity. During this shift, the color will move though every permutation of its particular hue. The similar term ‘color gradient’ refers to a shift from one color to another without respect to intensity. While these two terms have only superficial similarities, there is a great deal of confusion between the two and they are often exchanged inadvertently.
At its heart, an image gradient is as much a mathematical term as a graphics one. Originally, math was used to create the foundations for gradients before people were wholly capable of making them. These gradients are used as a method of determining vector direction and speed in color-shifting objects. The image-based gradient grew out of the applied uses of physics and became part of applied graphics.
When used in graphics, an image gradient is a description of the shift in intensity within a single color. Intensity is another graphics term that describes the amount of light reflected by a color. A high intensity means more of the color is being reflected towards the viewer, resulting in a color that is closer to white. A low intensity means the color is absorbing more of the light, resulting in a color closer to black. In either case, the underlying color is the same; it is just absorbing or reflecting more color.
In real terms, this creates an image gradient that looks like a shift from white to black through a single color. When applied to a standard picture, individual surfaces are covered in tiny gradients. In many cases, a single image will have color areas blended in order to make a smother picture. If the colors weren’t blended, there would be obvious artifacting in the image, usually resulting in a block-like appearance.
For instance, in a picture of a person’s face, there will be a gradient on the subject’s skin and a different gradient on her clothes. A computer can look at an individual image gradient and compare it to others covering a picture. Then it can find locations where different gradient areas touch one another. This information can then be used to find the edges of the subject in the image and derive a digital representation. This is often used as a method of correcting or enhancing images to bring out additional details.
