Understanding Recognition Types

Optical Character Recognition (OCR for short) is the overall conversion of typed or printed text into digital data.  OCR can recognize text from a variety of sources; scanned documents, a photo of a document, text in a photograph, or text overlaid on a photo (such a caption).  OCR works via a method called “pattern matching” where each letter appearing in an image or document is analyzed against a stored library of glyphs to determine what letter it is.  Once a match is made the OCR engine can then convert the image to machine-readable data.

Intelligent Character Recognition (ICR for short) is also known as “advanced OCR” in that it can accurately analyze and convert hand-written text to machine-readable data.  This gives it advantages over OCR because it is not limited to printed documents or certain fonts – in fact, its engines are constantly learning to recognize new handwriting patterns to improve its overall accuracy.

Optical Mark Recognition (OMR for short) is used as a method to interpret non-text markings on a document as data.  These markings can be supplied as answers to a test, votes on a ballot, or responses to a survey.  The ability to recognize markings is useful for enterprises as it allows them to limit the number or type of response to a specific element of data, such as a yes/no question on an application.  Once recognized, these markings can be fed into the processing system as useful business data.

Intelligent Document Recognition (IDR for short) is an umbrella term for the process of capturing, classifying, recognizing, and validating data from paper-based sources.  In IDR, documents are first identified by type and then customized templates are used to locate important business data.  Then, that found data is digitized.  This process is also known as “information capture” and is the core ability of the Captiva capture solution, offered by PaperFree.