The PDF is in this case a container for all the images. You can extract the images from the PDF with the tool 'pdfimages'. It is part of the poppler-utils. More info
here. Although it is Linux based, it might also exist for OSX.
The result are the separate images which can then be processed with ScanTailor.