Thread: OCR feature in Acrobat DC

When I run OCR on certain files, I get results that identify words like "Connecticut" as "comeeTicUt." Is there any way to manipulate a file to improve the accuracy of the output? I have tried managing this issue with the Optimize Scanned PDF option, but with limited success. I remember a time when Acrobat Pro had a feature that suggested fixes after running the OCR and before exporting the scanned content into MS Word.

Any ideas on what to do will be highly appreciated (before I run back to reinstall Acrobat Pro ).

Hi gentle!
I guess Acrobat DC also features this option. But since the interface has changed a lot maybe you will find hard to find it. Have you tried browsing the top navigation menu options?

I've found a similar feature under Tools>Enhance Scans to help you fix the recognized text. However, I took a sample PDF on-line from somebody who was asking about the same thing, and it doesn't seem to work the same.

Yes but the problem is the segmentation of the OCR. If it is a simple text, there is no problem but if we have a multiple designs it becomes complicated.

More after having tested it in a different file...It worked! You can use the Enhance Scans feature, but then you will go to Read text and the menu item at the bottom of the list that helps you fix the ocr'd text. It encircles in red all entries to be checked and you can start screening them one by one. It shows you the entry in the scanned copy and how Acrobat interpreted it, giving you the chance to make any necessary fixes. Once you approve an entry, it jumps straight to the next one, helping you go through all doubtful scans.

that's great, but the question is ... is it fast?

Originally Posted by RodrigoDG

that's great, but the question is ... is it fast?

Whether it may be fast or not will certainly depend on the quality of the input. Non-editable clear text is readily scanned, and just a few possibly misinterpreted words may show encircled in red and to be checked.

I tried a bad quality scanned copy in German and the outcome was not good at all! In any case, it is always an alternative to save the file in MS Word and run a spell checker to fix it.

Going back to your first post, I also guess the problem should be with some words that are not defined in the dictionary the OCR handles for the specific language you are scanning.

Originally Posted by gentle

Whether it may be fast or not will certainly depend on the quality of the input. Non-editable clear text is readily scanned, and just a few possibly misinterpreted words may show encircled in red and to be checked.

I tried a bad quality scanned copy in German and the outcome was not good at all! In any case, it is always an alternative to save the file in MS Word and run a spell checker to fix it.

Yes, it's an extra tool and helps to make easier conversions.