Gmail fights image spam with OCR

google.com

9 pointsbrlewis3 comments

That's pretty cool.

But there could be a reverse CAPTCHA where spammers send images with text that OCR engines can't recognize.

Any public breakthrough in OCR to solve the spam-CAPTCHA would actually hurt efforts to stop spammers who could beat CAPTCHAs to signup for more accounts. How ironic :)

Spammers have employed this technique since roughly about the time they started sending images. See here for some examples:

http://www.jgc.org/tsc.html

Also, real OCR is too expensive to do at the volumes of mail being sent today. Gmail is likely using an OCR-esque technique that has lower precision but is much faster to process a given image.

I don't suppose they use tesseract, though? I tried it recently, and it didn't recognize anything. Admittedly my test was not very thorough.

Reply on news.ycombinator.com