I’ve never been a huge fan of Optical Character Recognition (OCR) applications because of how they have a hard time reproducing the original document that is being used as the source. Sometimes, however, the formatting may not be as important as just trying to avoid retyping a few dozen pages that you only have in a printed form.
If you’re in that situation I’d definitely recommend checking out FreeOCR. It’s not the most extravagant OCR program I’ve seen, but it’s both free and really simple to use. Since it is capable of directly tying into most scanners I’d say that even a novice computer user could figure out how to use this.
You don’t have to use a scanner though. FreeOCR can also be used to open a PDF file or an image (including most common image types and TIFFs). I tested out the accuracy by printing a snippet from one of our articles, and then scanning it back in. The results I got were really good, but as of right now it can’t really recognize paragraphs. That falls under the formatting issues many OCR apps have since it puts a line break at the end of each line that it sees in the source. In the Text menu you’ll find an option to remove all of the line breaks, which is what I did in the screenshot above, and then manually put the line breaks in where I knew they needed to be. It’s not optimal, but it could sure beat having to retype a huge document.
Something to note is that they are currently running a private Beta test of FreeOCR version 4. This version, once released, will have even better accuracy and some kind of page analysis. It’s definitely worth keeping an eye on this project.
FreeOCR Homepage (Windows only; Freeware)