OCR and linux text and equations

On Tue, 31 Jan 2006 19:32:55 +0100, Horacio staggered into the Black Sun and said:

This is pretty normal. Math equations can be substantially more complex than English text. You have a lot more characters to recognize in simple math (English letters, bars, upper and lowercase Greek letters, arrows, infinity, fractions, set notation, maybe Hebrew letters...) than you do in English text. And instead of everything being left-to-right, some symbols can mean drastically different things if they're on top or on the bottom of a divisor. And then there are infinite series and integrals.

Kooka is a frontend for ocrad or gocr. gocr is better than ocrad. But no matter what, Linux OCR = teh suck. Sorry to say it but it does. Use Finereader if you have it; it'll give you much better quality than any Linux engine.

This is a difficult problem, and I don't think there's anything out there that does that right now. If you have math to enter, you'll probably have better luck doing it by hand with (La)TeX or LyX. If you're really ambitious, you can try to write something, but that'd be a long run for a short slide unless you had thousands of pages of math to OCR. (And you'd still have to proof it by hand afterwards!)

-- Matt GThere is no Darkness in Eternity-But only Light too dim for us to see Brainbench MVP for Linux Admin mail: TRAP + SPAN don't belong ----------------------------- penguins, is Tux." --MegaHAL