No way to correct OCR errors in Acrobat pro DC
There is a major problem with OCR in Acrobat Pro CD. I have recognized (OCR) text in many documents. They still contain a lot of recognition errors in the hidden text used for search. When I try to use the function/tool "Correct Recognized text" (translated from Norwegian), I get a message saying that there are no suspected errors. It is not possible to correct the remaining errors. This is just stupid, there is no 100% correct OCR program, there are lots of errors in the recognized text. There must be a way to fix the remaining errors, but your technical support tells me that it is not possible. Until this is fixed, I can not use the product.
Geir Thorud
shared this idea
-
Daniel Johnson
commented
I found there is a hidden way of doing this. In "Scan &OCR", click on "Correct recognized text." Acrobat may pop up a box with the message "Acrobat didn't find any text needing correction." If so, push "OK" to close that message box. Toggle on "Review recognized text." Notice the hidden text is now visible displayed over the image. Then somewhere in your text where an error exists, double click on it. A red box will appear over it. You can then correct it, by changing the text in the "Recognized as" text area. This option has some bugs. Sometimes it doesn't work, but if I close the dialog for "Correct recognized text" and again go through the process above, it usually works.
-
Farid Kerba commented
Scan and OCR recognition of text function has errors: I have just scanned an old text in pdf. But the endings of the lines, be it a point or a character letter or symbols, are not being checked, so when I highlight the text, it only does for the ones recognized. 2 options here, and I send you the file I have. You can try it yourselves and see what happens, or I have to reinstall Acrobat, or there is an actual problems bug happening.
-
Anonymous
commented
2 years later and still no fix for this issue...
-
Paul Dalton
commented
I'm using Acrobat Pro 2019. I'm having this same issue.
I've attached a copy of my original "scan to pdf" of a document (an old Trig table), as well as that same scan file after it has been "Optimized" by Acrobat.
I've also attached a Word document into which I copied/pasted several of the numbers from column 8 on the Logarithms page of the pdf file.
As the Word doc shows, each of these copied items has some OCR mistake(s) that should be corrected, but that does not get reported when "Correct Recognized Text" is selected.
Looking for a solution, I came upon this thread and I tried the recommendation of checking the "Check Review Recognize Text" option.
However, all that did was to show display a very faded view of just the image. I couldn't "do" anything with it, not even copy any of what appears to be text. Looking closely, the "text" in that faded image does not have any of the OCR errors that show up when "copy/paste" is used on the pdf file.
Isn't there SOME way to display and edit the content of the actual hidden text file?
Paul Dalton
-
Anonymous
commented
How the **** does "Check Review Recognize Text" solve the problem if you can't correct the ******* errors even if you can see them you ******* idiot
-
Hi,
There is also a functionality of Review Recognize Text.
You can enable it via
Tools -> Enhance Scans --> Recognize Text -->Correct Recognize Text --> Check Review Recognize Text.Can you please try this and let us know if this solves your problem.
Regards,
Gaurav Maheshwari
