Originally Posted by itimpi
That suggests that the original PDF file is an image based one, and not a text based one. In that case there is not much point in trying to convert it as the output is unlikely to be useful.
I agree. with itimpi's comment above. It would be possible to run the pdf through OCR, delete out the page images and convert that. It's a lot of work, but it will produce a smaller file than the original 7mb pdf and make the text reflowable as a decent ebook should.