PDF Tools·3 min read

How to Extract Images From a PDF at Full Resolution

A supplier sends a 60-page catalogue as a PDF, and you need twelve product photos for the website by this afternoon. Screenshotting each one gives you blurry crops at screen resolution. The photos inside the PDF are usually much sharper than what the viewer shows, and they can be saved as they are. This guide explains what an embedded image is, how the Extract Images from PDF tool finds them, which format to save in, and what to do when a picture you can see refuses to come out.

Open the free Extract Images from PDF All insights

What counts as an image inside a PDF

Not everything you see is a picture.

A PDF page is a set of drawing instructions. Some instructions draw text with a font, some draw lines and shapes, and some place a stored picture, called an image object, at a position and size. Only that last kind can be extracted, because only it exists as a grid of pixels inside the file.

Photos, scanned pages and most logos exported from design tools are image objects. Charts made in Excel, diagrams drawn in PowerPoint and logos saved as vector shapes are not; they are drawing instructions, and the only way to turn them into a picture file is to render the page.

How the tool pulls images out

The pixels as stored, not as displayed.

The tool opens the PDF with pdf.js in your browser and walks through the drawing instructions on each chosen page. Every time a page places an image object, the tool decodes it and saves it as its own file, at the pixel size stored in the PDF. A logo that repeats on every page is usually stored once, so it is saved once.

That stored size is often larger than the size you see on screen. A photo shown as a small thumbnail on a catalogue page may be stored at 2400 by 1600 pixels, and that is what you get. Each result shows its dimensions and file size under a preview, with its own Download button, and everything also arrives in one ZIP.

Choosing PNG, JPG or WebP

Match the format to the picture.

PNG is lossless and keeps transparency, so it suits logos, icons, diagrams and anything with a see-through background. It is also the safe default when you plan to edit the image later, because it adds no new compression.

JPG makes much smaller files for photographs, and the quality slider lets you trade size for detail. WebP is smaller again and loads quickly on websites, though some older desktop programs cannot open it. For a catalogue headed to a web shop, WebP or JPG at High quality is a sensible pick; for a logo headed to a designer, PNG.

A worked example: product photos from a catalogue

Twelve photos in two minutes.

Input: spring-catalogue.pdf, 60 pages, 38 MB. Drop it on the upload area, choose WebP as the output format, and type 12-23 in Pages, because the new range sits in that section. Leave Include small images unticked so the page icons and bullet graphics are skipped. Press Extract images.

Output: a grid of 24 images, two per page, each labelled with its page and size, such as p.14 image 2, 1800x1800. Download the twelve you need one by one, or take the ZIP named spring-catalogue-images.zip and sort the files on your computer. Nothing was uploaded at any point, which matters when the catalogue is not public yet.

When extraction finds nothing, or the wrong thing

Usually a vector drawing or a scan.

Most surprises come from a small set of causes.

No images found. The pictures are vector drawings, such as charts or shape-based logos. Switch What to convert to Pages to images and render the page instead.
One image per page on a scanned PDF. A scan stores each page as one big picture, so that picture is what you get. It is the original scan, which is often exactly what you need back.
A logo appears with a black or odd background. Some PDFs store transparency as a separate mask. Save as PNG, which keeps the transparency the tool applies.
Tiny files you did not want. Leave Include small images unticked, which skips anything under 50 pixels on both sides.
A photo looks cropped. The PDF may show only part of a larger stored image. Extraction gives you the whole stored picture, which can include parts the page hid.

Extract, render, or reuse responsibly

Two tools in one, and a note on rights.

The same page offers Pages to images, which draws whole pages at the DPI you choose. Use it for slides, previews and anything built from vector drawings. Use Extract images when you want the pictures themselves at their original quality. PDF to JPG and PDF to PNG open with the page mode selected and share every option.

Extracting a picture does not change who owns it. Supplier catalogues usually come with permission to use product photos; stock images in a report usually do not. Check before you publish anything you pulled from someone else's PDF.