Image are classified by the type according to their dimensionality and their representation:

• binary images: 1 bit/pixel. They are used for documents, and background/foreground maps
• grey-level images: 8 bit/pixel. These are used for outdoor scenes, and several industrial applications. This representation is currently a good compromise between information (most of the image information is carried by the intensity) and image size (ie, storage, processing, and transmission requirements).
• color images: 24 bit/pixel. These are used for multimedia applications. Usually 8 bits are reserved for each color component red, green, and blue.
• alpha-color images: 32 bit/pixel. These are also used in multimedia. They have an extra alpha "channel" (8 bits) that encodes the degree of transparency. A nice features of these is that 32 bits form a long word (usually an int) and so the image data can be aligned to word boundary.
• multispectral images: 8 N bit/pixel. These images are like "collections" of N grey-level images. They are used in remote sensing.
• sequence of images: 8 bit/pixel for each frame. These are used for remote surveillance.
• color video: 24 bit/pixel for each frame. These are used in multimedia applications.