At a fundamental level, digital images are just organized collections of numerical information. This information can be separated in varying ways called channels.

 

 As you probably already know, when an image is displayed on your computer screen, what you are seeing is actually a blend of red, green, and blue light. The values for these different colour components can be separated mathematically by computer software, and reproduced individually. Each one of these separate components can be called a channel.

 

You also know that we can think of an image in terms of its brightness and colour information. Photoshop can separate the numerical information from image pixels according to their brightness and colour. This method creates a Luminance channel for brightness information, and two chrominance channels, A and B, for colour information. This method of channel separations is used in the LAB colour standard.

 

In short, a channel is one of several possible streams of information that can combine to form an image.

 

Here is the Channels palette for an RGB image.

 

 

In the screenshot that follows, you can see multiple copies of the same image, with each copy displaying a different combination of channels.