Greetings.

A poor photocopy of photocopies with degraded readability; can A.I. in some sort of optical character recognition renew/retype text to a very readable condition ?

What would be the "instruction/request/text" to ask A.I. to

• perform such on a image with text, or;

• giving A.I. a .pdf document ?

Would it be something like "retype/reconstruct this document" attached ? (xxxxxxx.pdf)

Example with NO interest in the context but on its reconstruction and readability:

---

---> Can anyone teach me how to do it ?

I am not sure, Mastodon has a feature where if you upload a flyer for example, the alt text generator can take the image and try and extract the text elements (it has mixed results),, So I think the first stage may be just OCR and see what it comes up with,

If you can manually edit a few words, that may help, perhaps then run through a spell / grammar check or similar and see what comes from that.

If you want the output to be just readable, you don't want AI or anything inserting new words that could easily alter the meaning of the original text.

It probably requires manual intervention either way,