Extract PDF images

JPEG and JPEG2000 only. Those two are the formats whose raw stream bytes in a PDF are already a complete image file, so extracting them is a straight copy with no decoding involved. Most real photographs embedded in PDFs are JPEG, so in practice this covers the common case.

Because they are not in a format it can copy out safely. A flate-compressed bitmap would need its colour space and bit depth decoded and re-encoded into some container, which this tool does not attempt. Images whose stream runs through a filter chain are skipped too, since their bytes are still compressed and copying them would produce a file that is not a JPEG. The count of skipped images is reported.

You get one of two messages. If there were no embedded images at all, it says "no embedded images found". If there were images but none in a supported format, it names how many it found and says none are JPEG or JPEG2000. It never hands you a file with the wrong contents under a .jpg name.

No. The stream bytes are copied out exactly as they sit in the PDF, so what you get is the original image at its original resolution and quality. No generation loss, no resampling. If an image looks small, that is the size it was embedded at.

The PDF name, then the page number the image was found on, then the name the PDF gave the image object, with a .jpg or .jp2 extension. An image reused across several pages is written once, under the first page it appeared on, because objects are tracked by reference and repeats are skipped.

That one renders whole pages and gives you pictures of them, text included, at whatever resolution you pick. This one recovers the source photographs already inside the file, untouched. Use this to get a logo or a photo back; use the other to get a picture of the page.