Tesseract Ocr - Ui Automation 2019

Hi, i have some issue with the upgrade. I want use Tesseract OCR with Get Text OCR activity. But it doesn’t work i have this error . Can you help me please
image

Does that OCR work outside of CV activities?
Can you do a screen scrape with it?

@florinszilagyi

No it doesn’t work else if i use Screen Scraping.

Ah, so there’s a problem with the OCR engine.

Maybe Florin here can help, or maybe you could ask about it in the general section of the forum.

It seems there was a problem communicating with the Vision host. Do other OCR activities work?

We would also need a dump file to inspect why it’s failing, so if you’re able, we would appreciate if you could send one to us. To do this:

Create a new key (folder) in the registry, under Computer\HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\Windows Error Reporting\LocalDumps\ called UiPath.Vision.Host32.exe.

And create the following keys:

  1. DumpFolder (REG_EXPAND_SZ) – Folder where dump should be created (I recommend c:\temp but you can use any other folder where you have normal access, like the desktop or another partition)
  2. DumpCount (REG_DWORD) – Maximum dumps created before old dumps will be overwritten. The default is 10.
  3. DumpType (REG_DWORD) – Minidump (1) or Full dump (2). Use Full dump

You can see in the attached image how it should look like.

When the issue occurs again, try checking to see if there’s a dump file in your folder, and send it to us.

JUst a precision when it run once with screen scraping it works but when i use the engine manually it doesn’t work.
I can’t attach my dumpfile i send you in an email

The file is too heavy. How to send you? I can share it in my OneDrive

You can share it on a google drive or one drive.

It is done. This is the link
https://humans4helpsn-my.sharepoint.com/:u:/g/personal/h_diop_hub4it_com/Eag10hewT95GlpevcKryVLwBLwG9gTeY9TdapOkaYcGxng?e=xdHbZC

It seems Tesseract itself is failing.
Does this happen with any image you are trying to OCR or just with a specific set? If so, please send me the image it’s failing with.
Do you have any AV installed? Can you try disabling it and run the OCR again, see if it works.

Thanks

what is AV? It happens with any image only Microsoft is working .

Antivirus

No i haven’t installed an Antivirus

Hi,

no idea if it’s linked to same root cause, but on my side in UIPath Microsoft OCR is working perfectly but Tesseract OCR is failing systematically due to LoadEngine issue…
Appearing always after a full re-installation of UIPath Studio.

What language are you using for the language?
By default, only English comes with the activity, but you can add more languages (find a list here: https://github.com/tesseract-ocr/tessdata/tree/bf82613055ebc6e63d9e3b438a5c234bfd638c93)

Please check: C:\Users\<username>\.nuget\packages\uipath.vision\1.3.0\build\tessdata
Here, you should have at least “eng.traineddata” corresponding for the english language

If this does not solve the issue please attach the logs for more information.

Hi @florinszilagyi ,

thanks for your quick answer.
I’m using english language only.

I’ve created two topic for the issue including all logs.

Your help will be really appreciated !

Kind regards,
Quentin

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.