Computer vision exploits algorithms that allow computers to analyse images or multidimensional data drawn from the external world. The computer with its attached apparatus (e.g. camera, scanners) effectively serves as a device to “see” and “understand” visual objects according to the parameters set by the hardware and software. In so doing, computer vision allows us both to set up models that assist us in understanding processes of seeing in general and potentially to extract information that lies beyond the powers of the human eye and perception.