Term of the Moment

iconic


Look Up Another Term


Definition: OCR


(Optical Character Recognition) The machine recognition of printed characters. OCR systems can recognize many different fonts, including those designed specifically for optical recognition as well as typewriter and computer-printed characters. Advanced OCR systems can recognize hand printing.

From Bitmaps to ASCII
When a text document is scanned into the computer, a picture is taken of each page. Just like a digital photo, the page becomes a bitmapped image of pixels. OCR software then analyzes the light and dark pixels in order to recognize each letter and digit, which is converted to an ASCII character. See bitmap, ASCII file and pixel.

Hand printing is a lot more difficult to analyze than machine printing. Old, worn and smudged documents are also problematic. OCR is sometimes as much an art as it is a science.




OCR A Font
This is an example of the OCR A font. OCR A was designed specifically for optical recognition in the late 1960s when the average computer's processing power was dramatically less than it is today.






OCR Processing
When text documents are scanned, they are "photographed" and stored as pictures in the computer. OCR software converts the pictures into actual text characters, which take up considerably less room on disk.








OCR Machines
The "football field-long" machine (top) from Recognition Equipment, Inc. was used in the 1970s to process checks and credit card slips. The machine at the bottom is a contemporary unit. Both machines can handle OCR and MICR processing. (Images courtesy of BancTec, Inc.)