Software to extract original image files from PDFs

I'm looking for a program, free or paid, that can extract the original images from a PDF document.

You mean, you want the raw data?

PDF images are just that 'raw' pixel data in a two-dimensional array, along with some metadata about pixel representation.

You would need to save that raw data, the array dimensions, the pixel representation (i.e. grayscale, RGB, CMYK, …) and component width (1/2/4/6/16 bits). And perhaps PDF extensions provide additional variations.

The only file format that matches all that seems to be … PDF. (Or perhaps TIFF with all its various representations … )

I know there is a lot of programs that can export images, but all I've tried are actually creating a new image file and even give you the option to choose what format to save it in.

Difficult to say if any data manipulation takes place or not. Probably needs to do something like tiffdiff on an original image, a copy of which has passed through inclusion in PDF file, followed by extraction to 'same' format.

I'm looking for a program, free or paid, that can extract the original images from a PDF document. I know there is a lot of programs that can export images, but all I've tried are actually creating a new image file and even give you the option to choose what format to save it in. I'm actually hoping to extract the unconverted original file.

I'm not looking for this for any forensic purpose, it's actually data recovery related, so it doesn't need to be forensically sound.

Any suggestions?

As always happens, results will be depending on the source, the actual PDF creating software may have either "embedded" an image "as it is" or have transformed it to *something else* particularly losing resolution (in the case of bitmaps), etc.

I once had good luck with both Inkscape and XPDF (I had to take out the images related to some operating instructions in order to translate the text and all I was given was a .pdf)
https://inkscape.org/en/
http//foolabs.com/xpdf/
but those were "good" vector images, you can also try poppler (for Windows) pdfimages (still originated by XPDF that includes another version of pdfimages, but the poppler version should be more uptodate/recent)
http//blog.alivate.com.au/poppler-windows/

jaclaz

never mind, double post oops .

jaclaz

PDF images are just that 'raw' pixel data in a two-dimensional array, along with some metadata about pixel representation.

I'm pretty sure that's not always the case, though it might be with some PDF's. Just a few things I've observed with PDF's we recover that have bad sectors (so zero'd out sectors after recovery). Images that are embedded in a PDF that are a lossless uncompressed format will just show a gray box for the missing sector, while other files (I'm assuming which were jpegs before embedding) have the typical traits of a missing sector in a compressed image (where they're messed up from the bad sector on). Sometimes even in the same PDF we see images react those two different ways.

The reason I'm looking into this is that we have methods to repair many damaged files such as JPEGs which have missing sectors, but only if we have the original file. If it's converted at all after the corruption it's a new compression stream and we can't repair it.

As always happens, results will be depending on the source, the actual PDF creating software may have either "embedded" an image "as it is" or have transformed it to *something else* particularly losing resolution (in the case of bitmaps), etc.

I'm not at all concerned about conversion that happened when the file was embedded into the PDF, since the missing (bad) sectors happened to the PDF afterward. My concern is getting the image out of the PDF without converting it yet again. Thus maintaining the original compressed image data stream.