Patent For Google: Teaching Robots To Read
“Recognizing Text In Images” served as the main idea for a Google patent filed with the World Intellectual Property Organization.
The search engine could have another avenue for gathering information from the world it sees. A pair of Google’s engineers applied for a patent that would enable machines to read text included in pictures.
An InformationWeek report said the patent would allow tasks like searching videos by keywords appearing in the content, as one example.
“The method includes receiving an input of one or more image search terms and identifying keywords from the received one or more image search terms,” said the patent abstract.
“The method also includes searching a collection of keywords including keywords extracted from image text, retrieving an image associated with extracted image text corresponding to one or more of the image search terms, and presenting the image.”
Pulling off a technology that could do this would be quite a feat. Google’s patent goes beyond typical optical character recognition, where documents are scanned and read by software. In a real world full of myriad typefaces and languages, it seems Google will have a challenge in implementing the patent, should they have plans to do so.