THERE'S a hitch in Google's plan to digitise the world's books and make them searchable online: scanning them is taking too long.
That's because character recognition software needs a neat 2D image of the text. But book bindings cause pages to arch up either side of the spine - bending text and making it hard to interpret.
However, last week Google was granted a patent (US 7508978) on an answer to this problem. Its trick is to project an infrared pattern onto the open page spread. This lets a pair of infrared cameras map the three-dimensional shape of the pages by detecting distortion to the pattern. This in turn allows the distortion of the text to be determined - and therefore the degree of correction needed to read it accurately.
* Like what you've just read?
* Don't miss out on the latest content from New Scientist.
* Get 51 issues of New Scientist magazine plus unlimited access to the entire content of New Scientist online.
* Subscribe now and save
If you would like to reuse any content from New Scientist, either in print or online, please contact the syndication department first for permission. New Scientist does not own rights to photos, but there are a variety of licensing options available for use of articles and graphics we own the copyright to.
No comments:
Post a Comment