4. Automatic Data Capture (cont)

Optical Character Recognition (OCR)

Optical Character Recognition (OCR) enables the computer to identify written or printed characters.

An OCR system consists of a normal scanner and some special software. The scanner is used to scan the text from a document into the computer. The software then examines the page and extracts the text from it, storing it in a form that can be edited or processed by normal word processing software.

The ability to scan the characters accurately depends on how clear the writing is. Scanners have been improved to be able to read different styles and sizes of text as well as neat handwriting. Although they are often up to 95% accurate, any text scanned with OCR needs careful checking because some letters can be misread.

OCR is also used to automatically recognise postcodes on letters at sorting offices.

Speech Recognition

The user talks into a microphone. The computer 'listens' to the speaker, then translates that information to written words and phrases. It then displays the text on to the monitor.

This process happens immediately, so as you say the words, they appear on the screen. The software often needs some "training" in order for it to get used to your voice, but after that it is simple to use.