Is there any way to load model(tesseract custom model) at ones instead of loading every time?

[jitendra dubey]

We have pass every time model name while predicating the text from images. So I want to know any way to load tesseract custom generated model in "in-Memory" and predict faster.

Thanks in advance.

[Lorenzo Bolzani]

Yes, you can create an instance and reuse it. You call Init once and just reuse it. Performance does improve.

If you have multiple threads see this: https://stackoverflow.com/questions/4827924/is-tesseractan-ocr-engine-reentrant

For multi threading I created a pool of instances. On each instance I call init, setPageSegMode and setVariable only once, when I create the new instance.

Lorenzo

Il giorno mer 18 set 2019 alle ore 09:45 jitendra dubey <jitend...@gmail.com> ha scritto:

We have pass every time model name while predicating the text from images. So I want to know any way to load tesseract custom generated model in "in-Memory" and predict faster.

Thanks in advance.

--
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/0d8fb78f-e5a3-47d7-85a6-f8428062ea9b%40googlegroups.com.

[Zdenko Podobny]

You have to use tesseract API and not executable for this.

Zdenko

st 18. 9. 2019 o 9:45 jitendra dubey <jitend...@gmail.com> napísal(a):

We have pass every time model name while predicating the text from images. So I want to know any way to load tesseract custom generated model in "in-Memory" and predict faster.

Thanks in advance.

--