A PDF report, brochure, or scanned booklet often has real embedded photos inside it — but getting them back out as usable image files usually means an awkward screenshot-and-crop, which loses quality and adds a visible border or background every time.
Why extraction is different from screenshotting
A screenshot captures whatever’s rendered on your screen at whatever resolution your display happens to be — always a downgrade from the original. Extraction pulls the actual embedded image data out of the PDF file itself, at the resolution it was embedded at, with no re-rendering or recompression involved.
How it works
Extract Images scans every page for embedded photos:
- Upload your PDF.
- It searches the whole document for embedded JPEG images.
- Preview what was found.
- Download them individually, or as a single ZIP if there’s more than one.
What it can and can’t pull out
This reliably extracts JPEG images — the format most scanned and photo-heavy PDFs actually use internally, stored inside the file with a compression method called DCTDecode. That’s specifically what this tool looks for on every page, and it’s what makes a scan of a photo or a photo pasted into a report come back out clean.
It won’t pull out everything a PDF can contain, though:
- PNG-style embedded images — logos, screenshots, or diagrams that a PDF generator stored losslessly rather than as a JPEG — use a different internal encoding and aren’t picked up here. If your PDF is full of screenshots rather than photos, this tool may come back empty even though images are clearly visible on the page.
- Vector graphics — drawn shapes, logos, or icons built from paths rather than pixels — aren’t image files at all inside a PDF, so there’s nothing for an image extractor to pull out.
- Images nested inside a grouped object on the page (sometimes how design tools export a layered graphic) sit one level deeper than this tool looks, so they can get missed even though they’re technically embedded images.
If a PDF turns up nothing, one of those three is usually why — it’s worth opening the file and checking whether what looks like a “photo” is actually a screenshot, a logo, or a scan that was rendered as one flat page image rather than embedded as a separate photo.
Batch limits and password-protected files
This handles one PDF at a time — there’s no batch queue for a folder full of files, so a stack of reports means running each one through separately and downloading its ZIP before moving to the next.
A password-protected PDF won’t open here at all. Remove the password first with Unlock PDF — you’ll need to already know the password, since it doesn’t crack or guess one — then run the unlocked copy through this tool.
Nothing leaves your device
The whole extraction — opening the PDF, scanning its pages for embedded images, packaging them for download — runs in your browser tab as JavaScript. Your file is never uploaded to a server, which matters if what you’re pulling images out of is a scanned ID, a signed contract, or anything else you’d rather not hand to a third party just to get the photos back out.
The reverse direction
If you’re starting from photos and want to go the other way — building a PDF out of a set of images — JPG to PDF does that, combining however many images you add into a single PDF with one image per page.