Google and Book Scanning: 2004 Invention Discovered

May 3, 2009

Short honk: I enjoyed the “revelation” that National Public Radio (a “public” broadcasting entity that runs ads) discovered the secret of Google’s book scanning “machine” via an article in another publication. You can read the story here. The article points out that Google has a patent for this invention. There’s an image from the patent that shows one of Google’s helpful diagrams. A couple of points that merit a comment: First, the patent application was filed in September 2004. Assume that the Googlers Messrs. Lefevere and Saric started work on the project nine to 12 months before the application was filed. Google’s “machine” has been grinding away for six, maybe seven years. Hardly news unless one was ignoring the significant technical challenges Google had to overcome to make it possible for a small crew and part time workers to process documents in a wide range of form factors. Second, the reference patent US7508978 “Detection of Grooves in Scanned Images” is one component of a mosaic of inventions germane to the book scanning project. In short, there’s more to the Google scanning “machine” than one invention designed to correct for curvature of the source. You can explore a subset of Google’s patent documents here. Note that this set covers only 1998 to December 31, 2008, and is not a replacement for searching the USPTO collection.

Stephen Arnold, May 3, 2009


Comments are closed.

  • Archives

  • Recent Posts

  • Meta