Hi @Expiscornovus & team,
I am trying to use the If the text on Screen(OCR) in power automate desktop but not able to choose any OCR option from the drop-down, can anyone tell me if any settings have to be configured or enabled so I can use the OCR.
You need to use one of the Create OCR engine actions first, then just select the %OCREngine% variable: https://docs.microsoft.com/en-us/power-automate/desktop-flows/actions-reference/ocr
I have the same problem with not recognizing text with either free OCR engine. The text is clearly on the screen but no scenario I've tried has the OCR engine detecting the text. Is it possible that this functionality is only available with the paid-for license and not the free edition of PAD?
Same problem as @kelway here, same questions - is this functionality only available in the non-free version?
I've tried many scenarios with both Tesseract and MODI - read from the web, read from Notepad, read from image, etc.
Different behaviors for Tesseract and MODI, though. For Tesseract flows, no errors are generated if run from the editor, but when run from the Flows panel I always get this error:
The MODI engine always appears to be created, but extraction always fails with this error from the editor:
or this one (or similar) when run from the Flows panel.
For further clarification: following the steps from this PAD OCR video (Tesseract) generates the error message shown above.
UPDATE: MODI works. I belatedly realized there was a hoop to jump through: Install SharePoint Designer 2007 to be able to have MODI available. I just happened to have a copy of the installation file from years ago, but I don't know that it's readily available anymore. If you do find a copy (I'm sure you can if you try), follow the instructions here. Just FYI, I have the non-Office365 version of Office 2016 installed, 64-bit. The SharePoint Designer 2007 installed correctly, but I am anticipating problems with it if I ever update Office, as I believe it's 32-bit.
That minor potential hiccup aside, PAD has an activity (MODI OCR) that requires a Microsoft component that Microsoft no longer provides. That seems wrong.
Upgrage to newest version for PAD and You can perform the initialization directly through the actions that require the engines without using the Create Tesseract OCR engine action. (Only one action, We can get the OCR text now) as below:
More information: You can download the Traineddata Files for the specified language Tesseract's data here:
Success as below: (only one step)