scriptiii
100% Client-Side

Embedded raster only

Extract images from a PDF

Every embedded raster image, decoded at its stored resolution and offered as PNGs in one ZIP — with a preview so you can see what came out.

Extraction happens in this browser tab. Your document is never uploaded.

Short answer: drop the PDF and download the ZIP. Each entry is namedpage-N-image-M.png and holds the image at the resolution stored in the file. Vector artwork and page background colours are not images and will not appear.

What “embedded” means here

A PDF can store a photo three ways: as a JPEG, as a lossless raster with a mask, or as drawing instructions that merely look like a picture. Only the first two are images. That is why a page full of charts can yield nothing while the scan beside it yields one large PNG.

When a page render is the better answer

If you need the page as it looks — layout, text, vector logo and all — render it instead. Extraction gives you the pieces the file happens to contain, not a picture of the page.

Image extraction questions

Why do some PDFs give no images?

A page of text and vector drawings contains no raster image at all — the letters are drawing instructions, not pictures. A scanned page is the opposite: one big image per page, and those extract fine. If the tool finds nothing, the page really has nothing embedded.

Are the images at full resolution?

Yes. pdf.js decodes the stored pixels, so what you get is the image as it sits inside the file, not a screenshot of the page at screen resolution. A 300 dpi scan comes out at its 300 dpi pixel size.

Why are small images skipped?

Under 32 pixels on a side is almost always a bullet, a rule or a decorative corner. Including them would bury the two photos you actually wanted in a ZIP of forty slivers, and the result line reports how many were skipped.

What about a logo drawn as vectors?

It will not appear. A vector logo is a set of path instructions, not an embedded image, so there are no pixels to extract. Rendering the page to PNG is the tool for that job, not this one.

Related tools

PDF to JPGRender whole pages instead.Split PDFTake out the pages you need.Compress imageShrink what you extracted.