Yes - but I would have used it to start with, rather than type 500 pages from an image...
Look at OCR software - it looks at an image, and does the best job it can of converting that to text:
Google: open soiurce OCR[
^]
If you use a couple of them, and then compare all three text outputs you should have a fair confidence that most, if not all of the mistakes are caught.