Any idea if google is uing using http://code.google.com/p/ocropus/ to do this?

I was under the impression that was what they used for the book scanning project.