Share via

errors translating text from image embedded in docx...documents batch mode configuration

Rick Rinewalt 0 Reputation points
2025-09-15T15:28:27.3633333+00:00

am unable to get one successful docx with image translated correctly. tried multiple end-points with no luck.

Foundry Tools
Foundry Tools

Formerly known as Azure AI Services or Azure Cognitive Services is a unified collection of prebuilt AI capabilities within the Microsoft Foundry platform


1 answer

Sort by: Most helpful
  1. Azar 31,720 Reputation points MVP Volunteer Moderator
    2025-09-15T17:05:33.47+00:00

    Hi there Rick Rinewalt

    Thanks for using QandA platform

    Azure Translator itself only handles plain text, not OCR on images within Word files. If you run batch translation against a DOCX that has embedded pictures with text, the service will skip or error since there’s no text layer to extract. Tryy a two-step workflow: first run the DOCX through Azure AI Document Intelligence or Computer Vision OCR to extract the text from the embedded images, then pass that extracted text to Translator (or Translator Document if you need full document formatting). In batch mode, Translator only deals with text objects inside the document, so for images you’ll need preprocessing with OCR before translation.

    If this helps kindly accept the answer thanks much.

    Was this answer helpful?

    0 comments No comments

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.