Using my own detection instead of tesseract's

[Nima Afshar]

so i want to use my own detection alghorithm instead of the built in algorithm but i can't find anything helping me doing that in wiki or issues,does the api let me do that?

[Tom Morris]

On Monday, August 5, 2019 at 7:02:46 AM UTC-4, Nima Afshar wrote:

so i want to use my own detection alghorithm instead of the built in algorithm but i can't find anything helping me doing that in wiki or issues,does the api let me do that?

By "detection" do you mean page orientation detection? Layout detection? Something else? Being more specific about what you're trying to do will probably help folks give you better quality answers.

Tom  

[Nima Afshar]

By detection i mean text detection,by the way your right i should've been more specific thanks for advice tom.

Nima

[Timothy Snyder]

On my project I detect and crop down to textline level on my own. Then, with PSM 13, I give tesseract a single line of text.

--
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/8decc638-ed69-4dff-9aab-381fa8d3ec56%40googlegroups.com.