Getting the original photos back out of a PDF
The photos you need are in a brochure, a catalog or a report, and the original files are long gone. A screenshot gets you the picture at whatever size it happens to be on screen, cropped and with the page around it. A PDF usually keeps each picture as a file of its own inside it, often at a higher resolution than the page shows, and that stored file is what this tool saves.
Updated
Which tool you need
Photos from a brochure, report or catalog, as separate files
Extract Images from PDF. Every picture comes out once, in a ZIP, named by the first page it appears on.
A whole page as it looks, text and all
PDF to Image draws each page as a picture, which is a different job: everything on the page, at the resolution you pick.
A chart, logo or diagram built from lines and shapes
Shapes drawn by the PDF are not stored pictures, so there is nothing here to extract. PDF to PNG saves the page they are on, with sharp edges.
A scanned document, where each page is a photo of paper
Each scanned page is normally one stored picture, so this hands back the scans at the resolution they were made. Black-and-white scans saved as JBIG2 or fax-style images are the exception and cannot be read yet.
The PDF is too large because of its pictures
Compress PDF makes those pictures smaller inside the PDF itself.
Stored pictures and pictures of pages
A page in a PDF is a set of instructions: put this text here, draw this line, show that picture at this size. The picture is kept separately, at its own pixel size, and the page only says where to show it and how large. This tool reads those stored pictures directly and draws no pages at all.
So what you get can differ from what the page shows. A photo shown small comes out at its full stored size, and when a page shows only part of a picture, the file usually still holds all of it, which is what is saved. A picture repeated on many pages, such as a logo in every header, is saved once.
Why some come out as JPEG and others as PNG
Photos stored as JPEG are copied out exactly as they are stored, with no second compression and nothing lost. Every other picture the tool can read is decoded on your device and written as PNG, which is lossless and keeps any transparency the picture had.
A JPEG cannot carry transparency, so a JPEG photo that the PDF cuts out with a separate mask comes out as the full rectangle, and the result counts it. A photo prepared for print in CMYK stays CMYK when it is copied, and some apps display such files with strange colors.
What to know before you start
- JPEG 2000, JBIG2 and fax-style (CCITT) pictures, and pictures in spot colors, are counted in the result but not saved.
- Text and anything drawn as lines or shapes are not pictures and are not saved.
- A picture that has to be decoded and is over 50 megapixels, and a JPEG that would come out as a negative if copied, are counted and left out. A photo prepared for print in CMYK is not one of them: it is copied like any other JPEG. JPEGs are copied whatever their size.
- Past 2,000 pictures, the first 2,000 are saved and the rest counted. A PDF whose pictures add up to more than 512 MB is refused rather than saved in part.
- Password-protected PDFs need Unlock PDF before their pictures can be read.