Extracting embedded images from a PDF

Use pdfimages

pdfimages is a PDF image extractor tool which saves the images in a PDF file to PPM, PBM, JPEG or JPEG 2000 file(s) format.

It's a part of the poppler-utils package, which you'll need to install.

Usage: pdfimages [options] <PDF-file> <image-root>

  • option -all will extract images in original format.
  • option -j will extract images as .jpg (caveat: images are converted and usually size is larger than original)

Example1: The following extracts all images from a PDF file, saving them in their orginal format.

pdfimages -all in.pdf /tmp/out

Example2: The following extracts all images from a PDF file, saving them in JPEG format.

pdfimages -j in.pdf /tmp/out

Will save images from PDF file in.pdf in files /tmp/out-000.jpg (or /tmp/out-000.pbm; see below), /tmp/out-001.jpg, etc.

The pdfimages man page explains:

-j:  Normally, all images are written as PBM (for monochrome images) or PPM for
     non-monochrome images) files. With this option,  images in DCT format are
     saved as JPEG files. All non-DCT images are saved in PBM/PPM format as usual.

I often use Inkscape for this. Load the page, and delete all the other stuff. The advantage is that you can get vector images in SVG and modify them as you choose.


Note that this question is specifically asking about "Extracting embedded images from a PDF". The keyword is extracting! That means: I have a PDF; it has some images embedded within it; how do I get them out!? If that is your question, use pdfimages as the main answer by @pl1nk states.

How to convert a PDF into a bunch of images:

Many people Googling around and landing on this question (myself included), however, are searching for a slightly different question on not even realizing the difference until hours of frustration later. So, if you are looking for "How to convert a PDF into a bunch of images" instead, which is NOT the same thing as "how to extract images from a PDF", here's how: use pdftoppm. "PPM" here is an image format, so this simply means "PDF to image". It works extremely well, albeit slow for a modern multi-core system, since it's a single-threaded application and doesn't take advantage of multiple cores of processing power.

Ubuntu 18.04 comes with pdftoppm version 0.62.0. Check your version with pdftoppm -v:

$ pdftoppm -v
pdftoppm version 0.62.0
Copyright 2005-2017 The Poppler Developers - http://poppler.freedesktop.org
Copyright 1996-2011 Glyph & Cog, LLC

Read the manual pages with man pdftoppm to see all of its many useful features.

Supported output image formats:

As the man pages show, pdftoppm allows you to output images in the following formats:

  1. PPM (default)
  2. PNG (with -png)
  3. JPEG (with -jpeg)
  4. TIFF (with -tiff)

It also allows you to specify output in monochrome (-mono) or grayscale (-gray) (default is color), to specify page numbers, to place output images into a folder, to crop and resize, specify resolution, specify jpeg quality (between 0 and 100), specify TIFF compression, process only even or odd-numbered pages, etc. It works extremely well and is EXTREMELY USEFUL!

Here's some examples of how to use pdftoppm to convert a PDF to a bunch of image files:

  1. Output ppm files as pg-1.ppm, pg-2.ppm, pg-3.ppm, etc, in default 150 DPI x and y resolution:

    pdftoppm mypdf.pdf pg

  2. Same as 1, except place all of the output files in a folder called images:

    mkdir -p images && pdftoppm mypdf.pdf images/pg

  3. [My favorite] Output images into "images" folder in jpeg format with 300 DPI x & y resolution instead of the default 150 DPI. Note that the output images are at some default jpeg compression level, and will take up approximately 0.1~1 MB in space per file for 300 DPI resolution and assuming standard 8.5" x 11" PDF pages.

    mkdir -p images && pdftoppm -jpeg -r 300 mypdf.pdf images/pg

  4. Output images into "images" folder in jpeg format with 300 DPI x & y resolution, at the highest quality jpeg level possible! quality values can range from 0 to 100. See the man pages. With quality set to 100 and resolution set to 300 DPI, expect each jpeg file to take up 2x the storage as above, with sizes ranging from ~0.2~2MB, depending on the content, and assuming 8.5" x 11" PDF pages.

    mkdir -p images && pdftoppm -jpeg -jpegopt quality=100 -r 300 mypdf.pdf images/pg

  5. Output uncompressed .tif* images with **300 DPI x & y resolution. Output file sizes will be approximately 25 MB for 300 DPI and 8.5" x 11" PDF pages.

    mkdir -p images && pdftoppm -tiff -r 300 mypdf.pdf images/pg

Note that outputing each page above at 300 DPI takes 15~45 seconds on my slow computer, meaning that a 100 pg PDF could take as long as 100 x 45/60 = 75 minutes or so for 300 DPI jpeg images, for example.

To time how long the process takes on your computer, simply place the time command in front of the pdftoppm portion of any of the commands above. Ex: here's the output from converting a PDF which had 3 pages:

$ mkdir -p images && time pdftoppm -tiff -r 300 testpdf.pdf images/pg

real    1m47.572s
user    1m45.675s
sys 0m1.536s

This means it took a total real-life clock time of 1m47.572s, or 60 + ~48 = 108 sec, which is 108/3 = 36 seconds per page.

Related:

  1. [How to turn a PDF into a searchable PDF w/pdf2searchablepdf] How to turn a pdf into a text searchable pdf?
  2. How to turn a pdf into a text searchable pdf?
  3. How to convert PDF to Image?
  4. https://stackoverflow.com/questions/6605006/convert-pdf-to-image-with-high-resolution/58795684#58795684
  5. https://www.linuxuprising.com/2019/03/how-to-convert-pdf-to-image-png-jpeg.html
  6. How to programmatically determine DPI of images in PDF file?

Tags:

Pdf

Software Recommendation

Recent Posts