I’ve written before about how underwhelming the front page of Google books is, but the fact remains that they’ve got seven million books scanned, it’s a very useful service, and you have to wonder how they’ve actually done it? As with many things Google, there is a trick behind it.
A new patent reveals the process, which involves not just photographing (rather than scanning) the books, but building a 3D model of the pages using infrared light, so that the pages can be digitally visually flattened. This means they don’t have to go to the trouble of flattening each spread with glass, or pulling the books apart, both of which are time-consuming.
It’s another great example of Google using technology as a competitive inhibitor.
Now, how do they do the page flipping I wonder?
Check out the full patent if you want all the gory details.