A tool designed to deter fraudsters from registering fake e-mail accounts has been recruited to help digitise books and newspapers dating back hundreds of years.

Captchas are little boxes on web pages which show a squiggly set of letters and numbers that the user is required to transcribe correctly in order to register or enter the site.

They were devised eight years ago as a way of preventing computers from setting up e-mail accounts automatically which could then be used to send out spam, but a clever tweak means they are now being used to transcribe newspapers dating from the nineteenth century and earlier.

Instead of displaying a random collection of letters and numbers, the newly designed Captchas present the user with a word from an old manuscript that a computer, somewhere, is having trouble deciphering.

When three people type in the same word, the system deduces that this must be the one displayed on the manuscript, and relays this to the computer which has been stumped by the mystery word.

At present the system, which was devised by a 29-year-old assistant professor at Carnegie Mellon University, is partnered with the Internet Achive, a San Francisco-based not-for-profit organisation which is overseeing the digitisation of books at 70 public universities and libraries in the US - but it could conveivably be employed by any such project.

Luis von Ahn, who also created the original Captcha system, says that the new version - which is free to anyone who signs up and will be rolled out early next month - will be able to help digitise about 160 books a day.

(The original acronym stands for Completely Automated Public Turing Test test to tell Computer and Humans Apart, in honour of the pioneering British computer scientist Alan Turing.)

At present, mass digitisation projects such as the Internet Archive rely on humans to check over mistakes produced by a sophisticated scanning technlogy known as 'optical character recognition' (OCR).

OCR is used to translate the text on a scanned page into a more traditional digital format which can then be searched by users in a database, but the technology is not without its problems. For books published before 1900, for instance, the accuracy rate is estimated to be only 80 per cent

Mr von Ahn said he hoped the new system, called ReCaptcha, would enable his tool to be more productive.

"About 60 million Captchas are solved around the world every day - each taking roughly ten seconds," Mr Von Ahn said. "Individually that's not a lot of time, but in aggregate these puzzles consume more than 150,000 hours of work each day. What if we could make more use of this effort?"

About 45,000 sites - including Facebook and ITV - have begun employing ReCaptcha, Mr von Ahn said, but the number of participants is potentially limitless, and as more join, the speed at which books are being digitised will increase accordingly.

At present the number of sites involved means that the sections of a scanned book that are smudged, faded or otherwise unclear can be deciphered in an average of nine minutes. ReCaptcha also improved the accuracy of transcriptions to more than 99 per cent, Mr von Ahn said.

There are several current "mass digitisation" projects, including Google's, which is working with the Bodleian Library at Oxford, among others, and according to one estimate is digitising books at a rate of ten million per year.

A Google spokesman said the company used a mixture of its proprietary OCR software and humans as part of the digitisation process, but declined to say whether the project would partner with the ReCaptcha tool.

Explore Tech & Web

• Personal Tech

• The Web

• Gadgets & Gaming

Times recommends

• Point-and-go computers consign mouse to history
• YoYotech Fi7epower: the world's fastest PC
• Bridging the gap between mind and machine