Boneyard Tools

Extract Images From a PDF

Drop in a PDF and this tool walks every page, finds the embedded raster images each page paints, decodes their pixels, and lays them out as a thumbnail grid. Every thumbnail shows the true pixel dimensions and the PNG file size, and a Save button downloads that one picture. Everything runs in your browser with pdf.js, so the document and its images never leave your device.

How to extract images from a PDF

  1. Drag a PDF onto the drop zone, or click it to browse for a file up to 25 MB.
  2. Wait while the progress line scans each page, for example 'Scanning page 3 of 8'.
  3. Look over the grid of thumbnails, each labelled with its width by height and file size.
  4. Click Save under any thumbnail to download that single image as a PNG.
  5. Click Download all to save every image in turn, named image-p1-01.png, image-p1-02.png and so on.
  6. Click New PDF to clear the results and start over with another file.

Examples

Two photos on the first page of a report

A PDF whose page 1 paints two embedded photos
Thumbnails saved as image-p1-01.png and image-p1-02.png

A scanned document

A 3-page scan where each page is one full-page image
Three PNG files: image-p1-01.png, image-p2-01.png, image-p3-01.png

A text-only PDF

An invoice built from live text and ruled lines, no photos
The grid stays empty with a note that no raster images were found

Frequently asked questions

Does my PDF leave my browser?

No. The file is read and every image is decoded on your own device with pdf.js, the same open engine browsers use to show PDFs. Nothing is uploaded to a server, so confidential contracts, scans and reports stay private.

What kinds of images does it find?

It collects the embedded raster images each page paints, meaning photos, scans and other pixel pictures stored as image XObjects. Vector line art, gradients, background patterns and live text are drawn from instructions rather than pixels, so they are not extracted.

What file format are the images saved in?

Each image is decoded and re-encoded as a PNG, which is lossless and opens in every image viewer and editor. The label under each thumbnail shows the pixel dimensions and the resulting PNG size so you know what you are downloading before you save it.

How are the downloaded files named?

Names follow the pattern image-p<page>-<index>.png, where the page number is 1-based and the per-page index is zero-padded to two digits. The first two images on page 1 become image-p1-01.png and image-p1-02.png, which keeps them sorted in the right order in your downloads folder.

Why are there fewer images than I expected?

Identical repeated images, such as a logo watermark, are counted once per page they appear on rather than once per placement. Decorative shapes, gradients and vector diagrams never count as raster images, so a page that looks busy can still yield only a picture or two.

Can it pull images out of a scanned PDF?

Yes. A scanned page is usually stored as one large full-page image, so each scanned page gives you a single high-resolution PNG. That makes this a quick way to turn a scan back into individual page images.

Is there a file size or page limit?

The drop zone suggests keeping files under about 25 MB, and because decoding happens in your browser the real ceiling is your device memory. Very large or very high-resolution PDFs take longer to scan and use more memory while thumbnails are built.

Why did it say my PDF could not be read?

That message appears when the file is password protected or its structure is damaged, since pdf.js cannot open a document it cannot decrypt or parse. Remove the password in your PDF reader first, then try extracting again.

Does Download all give me a zip file?

No. There is no zip step, so Download all triggers each PNG as its own download with a short pause between them. Your browser may ask once for permission to save multiple files, which you can allow to receive the whole set.

Learn more

  • How images live inside a PDF

    How a PDF stores pictures as image XObjects, why vector art is different, and what a browser can and cannot pull back out of the file.

Related tools