Why would this be huge? There is a lot of previous work on recognizing text in images. Any reason why you think Google’s approach is technically superior to others? Have you talked to experts on text recognition about their technique?

Don’t get me wrong – there is a lot of potential in applying text recognition within Google’s business, so this may indeed be “huge”. But it is not clear that getting or not getting this patent will make much of a difference in their decision to implement this or not.