[Indic-computing-users] OCR and spell-checking for Asian scripts

[Sunil A. <su...@ma...>]

-------- Forwarded Message --------
From: Richard M. Stallman <rm...@gn...>
Reply-To: rm...@gn...
To: Sunil Abraham <su...@ap...>
Subject: OCR and spell-checking for Asian scripts
Date: Tue, 06 Sep 2005 13:33:31 -0400

Were you the one who was interested in the status of
spell-checking support and OCR for Asian scripts?
If it wasn't you, do you recall who it was?

I investigated this question, and found that
GNU aspell can handle nearly all of Unicode, provided
that a dictionary is available.  However, there are
specific problems with a few Asian scripts.
See http://aspell.net/man-html/Unsupported.html.

Some Asian languages already have dictionaries.
See http://aspell.net/man-html/Supported.html.

However, OCR is another matter.  The best free OCR program is GNU
Ocrad.  Its author says that each letter needs to be implemented by
hand--so that implementing a new alphabet is a lot of work.

He would be willing to act as a guide to whoever wants to
implement a new alphabet, but such a person has to be very capable
in order to do more than just waste his time.

--=20
Sunil Abraham, su...@ma... http://www.mahiti.org
314/1, 7th Cross, Domlur Bangalore - 560 071 Karnataka, INDIA
Ph/Fax: +91 80 51150580. Mob: (91) 9342201521
UK: (44) 02000000259