optical character recognition. An information processing technology that converts human readable data into another medium for computer input. An OCR peripheral device accepts a printed document as input, to identify the characters by their shape from the light that is reflected and creates an output disk file. For best results, the printed page must contain only characters of a type that are easily read by the OCR device and located on the page within certain margins. When choosing an OCR product, the prime consideration should be the program’s level of accuracy as it applies to the type of document to be scanned. Accuracy levels less than 97% are generally considered to be poor.