OCR Tech Allows Google to Index Millions of Scanned Documents
Scanned PDFs are a kind of darknet on a web — at best search engines see an image inside a PDF, but can’t parse out the actual text. But now that’s changed as Google recently announced that it will begin using OCR (optical character recognition) technology to index the text inside scanned PDF documents.
Although there’s no flashy new interface or anything tangibly different in Google’s search results page, the new technology means that the full text of the some 300 million PDF files in Google’s index will soon be converted to searchable text.
That’s quite a boost for your search results, though whether or not the PDFs show up in your searches depends a lot on what you search for. Google’s examples would seem to indicate that many of the these documents are very technical, like this guide to repairing aluminum wiring (follow the link and then click “view as HTML” to see what the results look like).
Lifehacker has a fairly novel way to put the new features to work for you — upload your scanned PDFs, tell Google about them with a link and then sit back and wait for your free OCR conversion.
Certainly there are faster ways of converting scanned documents and, given that most scanners ship with free OCR programs, we’re not sure how practical the idea is, but they get points for creativity.