Why don't you make a screenshot?
You can also load pdfs directly into a lot of image editors. I know photoshop can open pdfs and I'm almost certain Gimp (free) can too. That will allow you to grab the map at a larger size than if you screenshot it, and in PS you should be able to leave off the labels too.
Extracting images without transparency is easy, as Adobe Reader X (or Adobe Reader 8, though not 9) allow you to copy an image out of a PDF.
For the (often) used images with transparency (alpha channel) it becomes somewhat harder. I haven't found a cheap way (i.e. not investing heavily into Adobe products) to do that in a Windows environment.
But you can always install Ubuntu (with wubi that's not a problem or risk in any way) and use Evince (the preinstalled document viewer) to "save as" or simply drag and drop the images. With multiple images/layers it can be hard (if not impossible) to target the correct image, though. That's usually only a problem for title pages.
EDIT: Also, you'll get the (command line) tool "pdfimages" which (surprise) extracts images from a PDF (much like the afore mentioned product by SomePDF). It also doesn't handle transparency correctly (unless I've overlooked a parameter), but at least it extracts the alpha information as seperate pictures, so you can recombine them in an image editor like GIMP. Which btw. opens PDFs but creates a raster image of the whole page. Better than screenshots, though.
Ok, after posting I decided to do some further experiments, since it's been some time I've actively tried to solve that.
Opening a PDF in LibreOffice Draw (which can take a loooong time depending on the size of the thing, just press "OK" when it'd asking for the PDF password, it's not needed) results in some real quirky and bad pages. But the pictures can be saved just fine and seperate from the backgrounds including transparency.