Closed aunger closed 9 years ago
I've not much experience with OCR, but this sounds like it would be a great feature
Running the new single-screenshot of agent stats that the latest scanner produces and emails through the open-source Tesseract-OCR, all of the numerical statistics and enough of the labels came through cleanly. It didn't parse some of the separators or a few characters in my username, but those might be fixable with some training on the font.
Completed in tag 2.2
Proposed Scenario: John snaps a screenshot of his Agent page in Ingress app, scrolls down, snaps second screenfull. John switches to http://blueheronsresistance.com/stats in his browser, and taps Upload Screenshots button. John selects the screenshots from his camera roll and taps submit. Web app OCRs the text and records John's entry for the day.