I was wondering if there is some specific piece of software aimed at improving the text quality of .pdf or .djvu documents (or such then exported as images). I often have to deal with other peoples nasty scans of articles/papers or with scans of quite old (classic) papers/articles where the text quality is anyway not the best I have seen. I mean like some program with built-in "A.I." algorithms specifically for text recognition, rendering, reconstruction, (maybe some forensics tool?), text image optimization or similar. I have already had some partial successes with "levels" and sharpen filters in Photoshop, but those were rather the exceptions...
Thanks,
Marin
PS: I have attached a thumbnail of an excerpt so you can get the idea, and this one is in quite good shape. Click on the thumbnail to see it "full-size". Note also that the "full-size" is actually zoomed, so it appears quite bigger than the original (which would then be not as readable as here):





