What is OMR, ICR and OCR software, and how can it be of benefit?

All of these types of software are used for converting the scanned image of a paper document into information that can be processed using a computer. The sharper the scanned image, the more accurately the program will work.

A scanner treats a printed document as an image, which means when the document is scanned onto your computer it is static and nothing else can be done with it unless recognition software is used to convert parts of the document into something that can be used for the purpose intended.

Computer software that reads from scanned images of documents or forms, does so by recognizing black text or marks on the image and converting them into computer readable marks or characters.

The types of marks read by the computer is different on each of these optical recognition software types:

Optical Character Recognition (OCR), recognizes letters of typed or printed text.

Intelligent Character Recognition (ICR), recognizes letters of handwritten text.

Optical Mark Recognition (OMR), recognizes marks made in checkboxes on forms.

Optical Character Recognition

Brief History:

The practical uses of OCR became evident in the early 1950s, where research was done to solve the problem of converting printed messages received by the U.S. armed forces into machine readable messages for computer processing.

The first commercially used Optical Character Recognition machine was used by Readers Digest in 1955.

Overview:

When a typed or printed document is scanned it is saved as an image on a computer, and the text it contains can't be searched on or edited and neither can it be clearly printed.

OCR software converts the scanned images of documents, into word searchable documents which can be computer stored, edited or printed as required. The management of documents is one of it's most valuable uses these days, which replaces the need for paper based filing systems.

A real benefit of Optical Character Recognition software is the time saved on data entry of information from returned paper forms and the re-typing of text, such as from old books or documents, when there is no computer based original is available.

The Conversion Process:

The OCR software detects and extracts each character in the text of a scanned image, and using the ASCII code set, which is the American Standard Code for Information Interchange, converts it into a computer recognizable character. Once each character has been converted, the whole document is saved as an editable text document with a highest accuracy rate of 99.5 percent, although it is not always this accurate.

A high accuracy rate depends on the sharpness of the scanned image of the original document and whether a standard text font was used, which the software has already been trained to recognize. It does compare with the accuracy rate of data entry clerks but takes far less time to complete the process than it would to re-type documents or re-enter information from many forms into a computer system.

As a side benefit, text converted using OCR can be automatically translated into other languages and/or spoken during the process.

Intelligent Character Recognition

Overview:

ICR goes one step further than OCR and is trained to convert different styles of handwritten text (not cursive) into machine readable data. It uses specially designed forms containing blocks for the collection of hand written information. One character filled into each block for collecting data such as names and addresses and ID numbers from returned forms.

The Conversion Process:

This software allows different styles of hand printed text to be learned and converted into computer processable text. It is the least accurate of the three optical recognition types of software. The accuracy depends on the neatness of the hand printed characters, the
sharpness of the scanned form images and the repertoire of its learned characters.

ICR software often uses a system called a neural network to recognize handwritten characters and automatically updates it's database to remember new styles of handwritten characters. A high rate of accuracy for ICR is usually only 97 percent, and therefore requires editing of the captured data for perfect results.

Optical Mark Recognition

Brief History:

OMR hardware has been used since the 1960's typically for lottery tickets, but was implemented into the education system in the 1970's causing a revolution in multiple-choice assessment marking. In the mid 1990's the introduction of OMR software made this kind of technology more affordable and available commercially for all sizes of institutions and industries.

Overview:

OMR software analyzes multiple-choice hand marked responses from bubble sheet questionnaire forms. It extracts the determined marks and converts them into captured data information which can be used for automatic marking and reporting in education and market research etc.

Every type of OMR software uses it's own specially designed forms to locate where the checkboxes appear on the form and to identify who the form belongs often by using barcode identification or OCR.