How to Add an Alternate Description to PDF Figure Tags

Illustration of AI-generated alternate descriptions for PDF images using automated image recognition in PDFix.

Lucia Todova
Co-Founder | Chief Architect of PDFix Technology & Automation

Connect with Us on LinkedIn
Connect via email

Figures and images often contain critical information in a PDF – from charts and graphs to product photos and scanned documents. But without an alternate description, this content is invisible to people using screen readers. To comply with PDF/UA and accessibility requirements, every Figure tag should include an alternate description. In our latest video tutorial, we show you exactly how to add alt text in PDFix – both manually and automatically.

Why Alternate Desctiption Matters

  • Clarity & context – Well-written descriptions make complex charts or diagrams understandable.
  • Accessibility compliance – Required by PDF/UA, WCAG, Section 508, and EN 301 549.
  • Better user experience – Screen reader users can understand figures without relying on visuals.

Options in PDFix for Adding Alt Text to Figure Tags

In the video, we demonstrate several approaches:

1. Built-in Set Alternate Description

  • Enter a custom description manually.
  • Reuse text from closest tags (captions, headings, nearby paragraphs).
  • Extract text from behind the figure’s bounding box.

2. External AI Actions

PDFix supports external AI-driven actions for generating alt text automatically.

  • Choose different AI models, such as OpenAI or BLIP
  • AI suggests meaningful descriptions based on the figure’s content.
  • Review and refine results as needed.

3. Batch Processing

  • Select multiple files in Application View
  • Run a built-in or external action to add descriptions to all figures at once.
  • Filter which tags to handle (e.g., only Figures larger than 40×40 px).
  • Save results with custom output names or directories.

Continuous AI Model Integration

In PDFix, we don’t limit you to a single AI model for alternate descriptions.
We are continuously adding state-of-the-art AI engines to deliver the best possible results. Let us know your preferred model, and we can integrate it into the PDFix pipeline.

Watch the Video Tutorial

👉 How to Add an Alternate Description to PDF Figure Tags

The video walks you step by step through manual, AI-assisted, and batch workflows so you can ensure every figure in your PDFs is properly described.

Best Practices for Figure Alt Text

  • Keep descriptions short and meaningful.
  • Don’t duplicate captions – instead, complement them.
  • Skip purely decorative images.
  • Always review AI-generated alt text before publishing.

Watch the Webinar: Generate Figure Alt Text, MathML & Table Summaries Automatically

Learn how to automate alt text generation for figures, convert formulas to MathML, generate table summaries, and fix annotation descriptions – from basic templates to AI models.

Frequently Asked Questions

What does “figure has alt text” mean for PDF/UA compliance?

It means a validator checked whether a <Figure> tag has an Alt entry describing its content. Charts, graphs, photos, and scanned images all need this — without it, screen readers skip the figure entirely, which fails PDF/UA and WCAG checks.

What is a Figure tag in PDF/UA tagging?

A Figure tag marks an image, chart, or graphic as a distinct structural element in a tagged PDF, separate from surrounding text. PDF/UA requires every Figure tag to carry an Alt (alternate description) attribute so assistive technology can convey its content or meaning.

How do I add alt text to a PDF figure or image?

In PDFix, open the figure’s Tag Properties and either type a custom description, reuse text from a nearby caption or heading, or extract text from behind the figure’s bounding box. For images without obvious source text, an external AI action (OpenAI or BLIP) can generate a description automatically.

Can I generate figure alt text automatically with AI?

Yes. PDFix’s external AI actions analyze a figure’s visual content and suggest an alt text description — you can run this on a single figure or across an entire batch of files, then review and refine the results before finalizing.


Leave us a Question or Comment

Posted

in

,

Tags: