Optical Character Recognition (OCR) is a way to identify printed text, handwritten text, and numerical data and creating machine-encoded text by scanning in the information. It allows print to become electronic and easy to search, edit, store, display, and use in machine processing. Software identifies characters by program shape recognition and words by character proximity (Hosch). In the late 19th and early 20th century, the concept of optical character recognition was conceived in the idea to help blind people read (Schantz). In fact, a man named E. E. Fournier d’Albe aimed to create an instrument that would use the technology of the early “photophone” by Graham Bell to help those who couldn’t see by interpreting light to create various tones to be heard (d’Albe). Although this was the original idea, thought of in various ways by various people, optical character recognition has become a tremendous piece of technology used in our every-day lives.
A popular mark of history is in 1931 when US Patent 1838389 was given to Emanuel Goldberg for his “statistical machine.” The instrument was capable of reading text and translating it into Morse code, which was still commonly used at the time for various reasons. This patent later came into the possession of IBM (Schantz). Twenty years later, in 1951, David Hammond Shepard build a machine named Gismo, the first true optical character recognition system. Gismo had the ability to translate print into machine code for processing, could read aloud characters, and translate Morse code. In 1952, the Intelligent Machines Research Corporation was founded by Shepard and an associate, William Lawless, to commercialize their Gismo machine, later renamed the “Analyzing Reader” (Martin). In 1954, the first commercial optical character recognition machine was installed in a business—Reader’s Digest. In this time, the machine was used to automate and digitize business tasks, such as managing subscriber sales data and to convert the data into punched sales cards (Schantz; Smith). They later expanded their use of optical character recognition by using the system to digitize coupon serial numbers. Around 1965, the US Postal Service installed an OCR machine to sort mail more efficiently at a much faster rate (Smith).
Two fonts were created for OCR operations: OCR-A in 1966 and OCR-B in 1968. OCR-A was the first font to create a complete alphabet legible to both machines and humans alike (Smith). However, it was considered an ugly font that “wasn’t built for humans’ sake” (Smith). So OCR-B was designed by Adrian Frutiger (Smith). But in the same year OCR-A was developed, IBM had a breakthrough: the IBM 1287. This was a machining that possessed the ability to read handwritten letters. However, the two fonts were still used as the standard machine font for daily operations. In the 70’s, Ray Kurzwell met a blind man while traveling who explained his only real handicap was his inability to read text. Kurzwell claims this was the inspiration for his purpose for a project he was working on. This resulted in the Kurzwell Reading Machine in 1974, a machine that could take a page and scan it, parse the text, and recite the text to the individual (Smith). This was also the first OCR software capable of recognizing text in nearly any font (Hauger). It wasn’t until 1978 that these computer programs were sold commercially, and in 1980, Kurzwell Computer Products was sold to Xerox who renamed the software Scansoft and was able to further develop the technology (Smith).
Multiple innovations started to take place in the 80’s. The first passport scanner for the U.S. State Department was developed. American retailers began scanning price tags using OCR. Russian company ABBYY, founded by David Yang in 1989 and still exists to this day, began selling products to help digitalize paper files. Soon companies were created that offered online services and applications. Such an example is real-time foreign language translation through a smart phone app. Today, OCR is used for many things, even though most people don’t realize. Every time a file is scanned into a computer to be digitally stored, to edit, search, or hold, optical character recognition is used. From passport recognition and bank receipt data to assisted technology for those who are visually impaired to in-time language translations, optical character recognition continues to expand and redefine the software and recognition capabilities.
d’Albe, E. E. Fournier (1914-07-01). “On a Type-Reading Optophone”. Proceedings of the Royal Society of London A: Mathematical, Physical and Engineering Sciences. 90 (619): 373–375. doi:10.1098/rspa.1914.0061. ISSN 1364-5021. Accessed 11 March 2019.
Living in the early 20th century, E. E. Fournier d’Albe invented an instrument in 1914 that would help the seeing impaired to read. He believed that by using the same technology as the early telephone, his machine could read text and produce tones for the individual to hear and wrote to explain the instrument to other inventors and technicians of his time. Although this primary source contains very little information about OCR in general, it gives great detail about one of the first systems created, including a diagram drawing.
Britannica, The Editors of Encyclopaedia. “OCR.” Encyclopædia Britannica, Encyclopædia Britannica, Inc., 10 Nov. 2010, www.britannica.com/technology/OCR. Accessed 11 March 2019.
Martin, Douglas. “David H. Shepard, 84, Dies; Optical Reader Inventor.” The New York Times, The New York Times, 11 Dec. 2007, www.nytimes.com/2007/12/11/us/11shepard.html. Accessed 11 March 2019.
“Optical Character Recognition (OCR) – Computerphile : Computerphile : Free Download, Borrow, and Streaming.” Internet Archive, 4 July 2017, archive.org/details/youtube-ZNrteLp_SvY. Accessed 11 March 2019.
Schantz, Herbert F. The History of OCR, Optical Character Recognition. Recognition Technologies Users Association, 1982.
Herbert F. Schantz in an author and consultant for OCR and document processing. His writing is to inform students, scholars, and field-related experts alike, believing it to be a valuable piece of technology. He provides detailed information on the history and development of the idea and innovation of OCR machines and systems over the years, but his book his quite lengthy for a nice sit-down read.
Smith, Ernie. “OCR History: The Original Machine Learning.” Tedium, 22 Mar. 2017, tedium.co/2017/03/22/ocr-typography-optical-character-recognition-history/.
Ernie Smith is a blogger and journalist who has created it his personal mission to provide information on topics he believes people should be aware of but usually aren’t. He believes OCR is a piece of technology that is important to modern day society and how things are done. Most of the information in his article relates to the different fonts used for computers to read and the science behind it. However, the information he provides corresponds with other resources to give a good idea how OCR is used today.