Extract Images from PDF
Pull every embedded raster image out of a PDF (PNG/JPG) and download them as a ZIP — no upload.
Quick answer: Drop a PDF and we use pdf.js to walk every page's operator list, decode each XObject image, render it to a canvas and export as PNG. You can preview thumbnails per page and download images individually or as one ZIP — all in the browser.
Last updated
Frequently asked questions
- What kind of images does it extract?
- Embedded raster images (photos, scanned pages, logos, charts saved as bitmaps). Vector graphics drawn with PDF path operators are not images and can't be extracted as PNG/JPG — they live as drawing instructions.
- Why does my PDF show 0 images?
- Two common reasons: 1) the PDF is text + vectors only (no embedded raster images), or 2) graphics are stored as inline images that some PDFs encode in a way pdf.js can't surface. Try converting the page to PNG with PDF to PNG instead.
- Are the images uploaded?
- No — pdf.js parses your PDF and the canvas API encodes the bitmaps locally. Open DevTools → Network and confirm: no image data ever leaves your browser.
- Will the images keep their original resolution?
- Yes — we extract at the embedded pixel dimensions. If a logo is stored as 200×50 px in the PDF, you get a 200×50 px PNG, not a re-rendered upscaled version.
- Why are some images PNG instead of JPG?
- We export everything as PNG to keep extraction lossless — re-encoding a JPG to JPG would add a second compression pass. If you want JPG, run the result through Convert to JPG.
- Does it preserve transparency?
- Yes — PNG output keeps the alpha channel, so logos with transparent backgrounds stay transparent.
- Can I extract from a password-protected PDF?
- Not directly — unlock it first with Unlock PDF, then extract images from the unlocked file.
- How big a PDF can I throw at it?
- Up to a few hundred pages with hundreds of images is fine on a modern laptop. Memory is the limit since each image is decoded into a canvas; very long catalogues with thousands of images may need to be split first.
- Are duplicates deduplicated?
- Yes within a single page (same XObject reference), but if the same image is embedded twice as separate objects, you'll see both — that's how it lives in the file.
- Why do scanned PDFs show one big image per page?
- Because that's what they are — a scanned PDF is usually one full-page bitmap per page. To get a tighter crop, run our Crop PDF first or use PDF OCR to extract text instead of images.