As reported yesterday, Google Docs API started testing a new text recognition feature, as in Evernote. That is, it will be possible to simply take a picture or scan a document, upload a picture to Google Docs - and the document will appear there as normal text.
To activate OCR when loading documents via the API, you need to add the parameter ocr=true .
A working demo version appeared online (the sources will be later). JPG, GIF and PNG files of up to 10 MB are supported. Recognizing one page of text takes up to 40 seconds (today, obviously, it will take longer). The service works with glitches, errors are possible. ')