Hi everyone,
I’m working with the new OCR Image block to extract text from images, but I’m encountering challenges with image input methods. The Quick Help section doesn’t currently have documentation for this block, so I’m hoping the community can provide guidance.
What I’m Trying to Accomplish:
Extract text from PDF images using the OCR Image block in a secure workflow for a client project.
Methods Attempted:
I’ve tested several approaches to provide images to the OCR Image block’s Image URL field:
User Input Block (Upload File, No Processing) → assigned resulting variable to Image URL field
Google Workspace upload → copied link directly to Image URL field
Google Workspace upload → copied link with {{variable}} syntax in Image URL field
Current Results:
All three methods result in “Model Error” messages in the debugger.
My Working Theory:
Based on troubleshooting, it appears the OCR Image block may require a direct, publicly accessible URL (similar to other MindStudio image processing blocks).
Security Concern:
This workflow handles sensitive client documents that must remain secure. If public URLs are required, are there any secure alternatives for creating accessible image links that don’t compromise document confidentiality?
Questions for the Community:
Has anyone successfully implemented the OCR Image block with secure file handling?
What’s the proper method for providing images to this block?
Are there secure URL generation options that work with MindStudio’s image processing requirements?
Any guidance or documentation references would be greatly appreciated!