Page 1 of 1

ABBY FineReader (OCR program) - teaching it polytonic Greek

Posted: December 31st, 2015, 6:08 pm
by griroz
Hi.
ABBY FineReader can be taught to recognise any alphabet. Normally it can recognise only modern Greek characters, but it can be taught to recognise any Unicode characters.
I tried it once but my approach must have not been the best to teach it polytonic Greek.

Has anyone already done it?
If so, would anyone be able to share the file of already learned Ancient Greek "language" or share the file(s) (images) he/she taught FineReader with it? I don't know how many letters should I give to it and how many different files, to make it "learn".

If anyone, thanks

Re: ABBY FineReader (OCR program) - teaching it polytonic Gr

Posted: January 2nd, 2016, 10:06 am
by Jonathan Robie
The best OCR results I've seen for older volumes of polytonic Greek come from Bruce Robertson's Lace project (http://heml.mta.ca/lace/), which is open source. There are classifiers for different font types, so it helps to be able to identify the font.

For more recently printed works, Nick White's Ancient Greek OCR project (http://ancientgreekocr.org/) may give better results, according to rumor.

I don't know of similarly good training materials for ABBY FineReader, but I don't have that program. Bruce and Nick are probably the world's leading experts on ancient Greek OCR, they may be able to tell you more.

Re: ABBY FineReader (OCR program) - teaching it polytonic Gr

Posted: January 2nd, 2016, 1:36 pm
by griroz
Thank you very much for the contacts! :)

Re: ABBY FineReader (OCR program) - teaching it polytonic Gr

Posted: January 3rd, 2016, 8:04 am
by Jonathan Robie
griroz wrote:Thank you very much for the contacts! :)
You're welcome!

And one plea: if you are scanning in old texts, please consider (1) licensing them under a Creative Commons license and (2) making the text available on Github so people can easily get the source text and offer corrections.

Re: ABBY FineReader (OCR program) - teaching it polytonic Gr

Posted: January 3rd, 2016, 8:44 am
by Ken M. Penner
griroz wrote:If so, would anyone be able to share the file of already learned Ancient Greek "language" or share the file(s) (images) he/she taught FineReader with it?
I have done some training for ABBYY FineReader 12 on the Athenaze textbook (second edition, I believe). The user pattern and language file can be accessed at https://onedrive.live.com/redir?resid=9 ... t=file.fbt