I'd be rather wary of trying to get useful results by applying OCR to a scanned
academic book -- academic books tend to have things like mathematical or other technical symbols, tables and charts, etc., none of which will survive OCR very well. Unless you're willing to reconstruct all that stuff manually, I'd keep it image based. (Unless it just happens to be all straight text, in which case, go ahead and give it a try.)
The question is:
are the pages upside down or sideways in the scan? How does the file open in Adobe reader, for example? Do you need to rotate it there to read it on your screen? If so, there are lots of tools to turn them around the right way.
pdftk comes to mind, though there are others.
I second the recommendation for
PaperCrop and especially
pdflrf, but even those might not work properly until the pages in the scan are oriented the right way.