I'm fairly certain that PDFBox has no provisions for extracting images from PDFs. Text - yes, images - no. A quick look at its feature list would seem to confirm that.
Edit: I take that back - it seems that PDFBox does have some fairly low-level routines for image extraction which do work for some kinds of images, although not all of them. So it would seem to depend on the PDF in question, and its embedded images,like James said.