↑↓ to navigate Enter to open Esc to close

Extract PDF Images: Deduped, Honest About Quality

A repeated logo is extracted once, not once per page. Identical images are recognised by content and merged into a single result showing every page it appeared on. Every other tool checked pulls it out again on every page it finds it.

Pulls the actual embedded pictures out of a PDF — not a screenshot of the page, the original image object. Nothing is uploaded and there is no limit.

Extract PDF Images

Local
Choose a PDF, or drop it here Nothing leaves this tab
Decoded by PDF.js, fetched when you add a file. Your document is not part of that request.

An embedded image, not a screenshot of the page

A PDF page can hold text, vector shapes, and separate embedded image objects — a photo, a logo, a scanned signature — each stored as its own object inside the file. Extracting images means finding those objects and pulling them out; it is a different operation from photographing the whole page, which is what a page-to-image converter does instead:

ToolWhat it producesUse it when
This pageThe original embedded image objects, one file per pictureYou want the photo, logo or scan that is inside the PDF
PDF to ImagesA picture of each whole page, text and allYou want a thumbnail, a preview, or a page to drop into a slide

A page rendered as an image includes everything on it — margins, text, headers. An extracted image is only the picture itself, at the size it was placed in the document.

Why this is decode-then-re-encode, and what that costs

A browser has no built-in way to reach into a PDF and copy an image's raw bytes untouched. What it can do is decode the image the same way it decodes one to display it — and that decoded picture is what gets written out, as PNG:

What happensWhat it means for you
DecodingThe PDF engine turns the embedded image into ordinary pixels, the same step it takes before drawing the page on screen
Re-encoding as PNGLossless from those decoded pixels — nothing is thrown away at this step
Not byte-identicalThe output file is not the same sequence of bytes the PDF stored, even though the picture is visually the same
File sizePNG is usually larger than the original encoding, particularly for photographs
"Original quality, no compression" is not quite true anywhere in a browser. Every tool checked makes this claim without qualification. What is actually happening, here and almost everywhere else, is a decode followed by a lossless re-encode — visually identical, not byte-identical. The distinction rarely matters; it is worth knowing before assuming the output file is a copy of what the PDF stored.
A logo embedded once but placed on many pages of a PDF is recognised by its content and merged into a single extracted result, rather than extracted again on every page it appears on A REPEATED LOGO, RECOGNISED ONCE page 1 page 8 page 22 one result found on 3 pages others: three separate files, same picture Every tool checked would hand back three identical files here. This one hands back one, and says where it was found.
Recognising the repeat needs nothing clever — the extracted image is fingerprinted and a second identical fingerprint is merged rather than added again.

When an image cannot be pulled out

Almost every embedded image decodes cleanly. A small number do not, usually because of how a particular scanner or export tool encoded them:

SituationWhat you see here
Ordinary photo or logoDecodes and extracts normally — the overwhelming majority of PDFs
An unusual scanner encodingFlagged in the results as could not be decoded, on the page it was found, rather than skipped silently
An image reused with a different crop or tintExtracted as a separate result — the content is genuinely different, so it is not merged with the original

A failure is reported, not hidden. A tool that silently drops an image you needed is worse than one that tells you it could not get it.

Common mistakes

Using this to get a picture of a whole page. This pulls out embedded image objects only — text and layout are not included. For a picture of the page itself, use PDF to Images instead.
Expecting a smaller file than the original embedded image. PNG is lossless, and lossless is usually larger than whatever compressed format the PDF used originally — particularly for photographs. If the picture is a photo and file size matters, re-compress the PNG afterwards.
Uploading a confidential document just to pull out one image. Contracts, ID scans and internal reports are exactly what people extract images from, and exactly what should not sit on a server whose retention policy you are trusting. Local extraction removes the question.

Frequently asked questions

Is my file uploaded?

No. It is read and the images are extracted entirely in your browser — the network panel in developer tools shows no request carrying your document. The only network request is for the PDF engine, and it happens once.

Why are the images PNG rather than the original format?

A browser has no way to copy an embedded image's raw bytes untouched from inside a PDF — it can only decode the image the way it decodes one for display, then write out what it decoded. PNG is the lossless format for that: nothing is thrown away in the re-encode, though the output is not byte-identical to what the PDF stored.

Why does one image show up only once when it was on several pages?

Because it is the same image, recognised by its content rather than extracted fresh on every page. A letterhead logo repeated across forty pages becomes one result here, with a badge showing how many pages carried it — not forty separate identical files.

What if an image cannot be extracted?

It is reported, not skipped silently. A small number of images use encodings that will not decode cleanly in a browser; when that happens the page it was found on is shown with a note, so you know something was there rather than assuming the document had no images.

Can I extract images from just some pages?

Yes. Type a range like 1-5, 8, 12- in the pages field before scanning, or leave it as all to check the whole document.

Does this get a picture of the whole page?

No — only the embedded image objects inside the page, not the text or layout around them. For a picture of an entire page, use PDF to Images.

Why is the extracted file bigger than I expected?

PNG is lossless, which for a photograph is usually larger than the compressed format the PDF originally used. The picture has not gained detail; the file format is simply less aggressive about discarding it.

Is there a page or image limit?

No cap is imposed. Nothing runs on a server, so there is nothing to ration; the practical ceiling is your own machine's memory on a very large document.

Related tools

See the full list of Document Tools, or try: