Batch PDF Image Alt Text

About Batch PDF Image Alt Text

When an accessibility check reports dozens of images without alternate text, opening each one in an editor is slow. This tool lays out every image in the document as a thumbnail grid so you can write all the descriptions on one screen.

Each description is saved as the /Alt entry of a Figure element in the tag tree (the Figure that already tags the image, or a new one). If the file has no tag tree at all, the tool first creates a minimal structure root and sets the Marked flag in MarkInfo.

Good alt text says what the image is for

A screen-reader user needs to know what the picture contributes at that point in the text. "Chart" is not enough; "Quarterly revenue 2024, highest in Q4" gives the takeaway. A logo only needs "Acme logo". Borders and background patterns should be left blank, and blank entries are skipped.

An image reused on many pages appears once in the grid with its page numbers. For a header logo repeated on every page, leaving it undescribed spares listeners the same phrase on each page.

How the Figure tag is tied to the actual image

For a Figure tag to point at the actual image, the drawing operators in the content stream must be wrapped in BDC/EMC and linked by an MCID. This tool wraps each operator that draws the image (Do) in “/Figure <</MCID n>> BDC … EMC”, records the page and MCID in the Figure element, and registers “MCID n on this page → this Figure” in the structure tree's ParentTree. When the same image appears on several pages, one Figure points at every placement. That is how PAC and Acrobat's full check know which picture a description belongs to.

An image already tagged as a Figure gets no new element; only that Figure's /Alt changes. Its current description is prefilled in the grid, and applying the tool twice to the same file doesn't add Figures. Images marked as artifacts or drawn inside another tag, such as a paragraph, are left alone and the number skipped is reported. New Figures are appended at the end of the structure tree, so if reading order matters, move them into place with the PDF tag tree editor.

Text inside images and complex charts

If an image contains words, such as a slogan or a chart title set as a picture, copy those words into the alt text; screen readers cannot read text that is part of an image.

Listing all twenty values of a bar chart in alt text is hard to follow by ear. Give the trend and the conclusion in the alt text and keep the figures in the body text or a table. Grid thumbnails are small, so check chart values on the original page before you type.

Worked examples

How to use it

1. Open the PDF; every image appears in a grid.
2. Type alt text only for images that carry meaning.
3. Apply, then re-run the PDF accessibility checker.

Frequently asked questions