Existing OCR scripts in JavaScript [closed]

Question

I have an idea for a CMS enhancement, to extract text information from images (for example, scanned documents), and want to know if there is already anything out there to help me along?

Basically, I want to know if there is an existing OCR script written in JavaScript that can extract sentences/words from an image (using canvas, for example).

I know there are some scripts that do relatively small tasks such as captcha-cracking, but I haven't yet come across a script for extracting full sentences.

Is there such a thing, or would I need to write it from scratch?

Answer 1

Antimatter15's Ocrad.js is a possibility