How to use tesseract 4 engineMode 2 ( Legacy + LSTM engines)?

3,669 views
Skip to first unread message

于洋

unread,
Jul 13, 2018, 12:01:18 AM7/13/18
to tesseract-ocr
Tesseract 4 introduced new LSTM engine. The LSTM engine needs LSTM trained data, and the legacy engine needs old trained data. Two types of trained data are incompatiable with each other.

When I set OCR Engine to 2, it will use Legacy and LSTM engines. But how can i provide two types(LSTM and legacy) trained data for tesseract?
OCR Engine modes:
  0    Legacy engine only.
  1    Neural nets LSTM engine only.
  2    Legacy + LSTM engines.
  3    Default, based on what is available.

Shree Devi Kumar

unread,
Jul 13, 2018, 12:04:29 AM7/13/18
to tesser...@googlegroups.com
The traineddata files can hold both types of models. The OCR Engine mode chooses which ones get used.


--
You received this message because you are subscribed to the Google Groups "tesseract-ocr" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tesseract-oc...@googlegroups.com.
To post to this group, send email to tesser...@googlegroups.com.
Visit this group at https://groups.google.com/group/tesseract-ocr.
To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/bceef2a8-4f45-4372-8392-78679110f8b5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


--

____________________________________________________________
भजन - कीर्तन - आरती @ http://bhajans.ramparivar.com

于洋

unread,
Jul 13, 2018, 4:41:39 AM7/13/18
to tesseract-ocr
Thanks a lot.

I have combined LSTM and legacy together. Using combine_tessdata command.
Reply all
Reply to author
Forward
0 new messages