Bold italics detection in tesseract 4

[Ravi Nemala]

I read tesseract 4.0 does not support wordFontAttributes in LSTM mode. Can I just use oem 0(Legacy) to identify the bold/italics in my image?

[Zdenko Podobny]

I do not have 3.5 version available, but simple of test image with tesseract v5.0.0-alpha-479-g247c show "some detection" of bold but on wrong places. So I would not suggest to use >=4.x version for this tasks.

Zdenko

ut 15. 10. 2019 o 16:52 Ravi Nemala <ravi....@gmail.com> napísal(a):

I read tesseract 4.0 does not support wordFontAttributes in LSTM mode. Can I just use oem 0(Legacy) to identify the bold/italics in my image?

--
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 view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/5d720ed7-2dd4-453d-98da-087c13770b31%40googlegroups.com.