Garrett, did you scan the text with the scanner software explicitly set up for scanning documents? If it's set to scan color or black and white (grayscale) photographs, the scanner will pick up a lot of shadowing and whatnot, and really pisses off the OCR software, because it'll think various shadows and whatnot are some sort of text or images. If it's set up correctly, all your scanned images should be duotone--no shades of gray. It makes it a hell of a lot easier for OCR software to do its thing. If the text is relatively clear, there shouldn't be that many typos popping up when the images are run through the OCR software.