ABBYY FineReader PDF

Easy of usage, processing batch files, a lot of different output format, standard OCR capabilities, and opening to OCR custom training

Output to Excel (tables are really not fully detected, even text within cells, with a lot of work to do in the end) and Word (every text appear converted in annoying frames, difficult to remove or manage at all) mostly have to be really improved to be the best.
OCR must be improved very much. You have bad result if you want to recognize document in Greek (most part of diacritics are not well detected) and, with Hebrew, is literally a worst situation (you have not vowels and all diacritics). If you try to do by yourself, using custom OCR training, it's a very long journey and you don't even get a good result, and many times it's even not so clear if you are using your custom training set or not and even is really not clear why you cannot use one only feature to perform "verify" and "ocr training" in one only function/feature/action.

That only sometimes renders a perfect conversion into Ms word that I can use to work, and I mean a file where I don't have to remake any parts of the text or recapture images, etc.

That it changes the page I'm working on just by scrolling, that's tedious. That's solves many of the text by putting it into a floating case box, and that's terrible and cannot be turned off. That many obvious tables are not detected by Abbyy when automatic converting, specially those without visible lines. That it fails to detect structures when converting scanned pages. That the function of detecting the headers and foots almost always fails, but instead it tries to convert as headers parts of the text that are not. Correcting all the above is time-consuming, which is contrary to the purpose of this program. And recently there is a contradiction between doing things in the main window of the OCR editor I'm doing them at the detailed window, for example if I try to select one cell to be treated as an image, it's conflicts the problem if I chose it at the main window or at the detail window, with the first one generally overriding the second. I'm now in Windows 11 the program crashes a lot.

Once I familiarized myself with the interface, the program worked as needed. No need to keep referring back to Dr Google or support.

The initial interface. It took me a few weeks to look for things. For example, it was easy to delete a page...remove a page and set up a different PDF? That one has too many steps.