Face detection is the process of scanning an image in order to identify human faces within the context of other objects or a background. There are several algorithms that perform this task, but most involve a computerized classifier making a binary decision about an image as to whether it is or is not a human face. Typically, the classifier examines all areas of an image by breaking it up into equal pieces and comparing the pattern in each piece against patterns that it has previously learned match human faces. This process is then repeated at a different scale of equal pieces, and then repeated again and again at all scales, until the scale is so small (a picture taken from a distance at a stadium crowd) or so large (extreme close-up of half of the subject's face) that facial patterns can't be recognized. Modern hardware can perform this classification very rapidly, many cycles per second.