You are viewing limited content. For full access, please sign in.

Question

Question

Incorrect OCR metadata

asked on May 11, 2015

Hello All,

I am having a case where the OCR'd pages column reads "All" but within the document, not all the pages are OCR'd:

Is there a way to "UN-OCR" a batch of entries? I imagine that would involve deleting all the txt files in the repository. If the txt file is removed, will the client recognize that the OCR setting has changed?

I would prefer to mark them all as "some", which is accurate. In this case, there are 66K+ documents that 'might' be OCR'd. Reprocessing all of these will take months. Any suggestions?

 

Thank you in advance!

 

0 0

Replies

replied on May 12, 2015

 

Hello Chris

According to development, we need first to confirm whether the documents can be ocred or not, by this we mean Open Laserfiche Client, locate any of the documents in question, select the document right click choose option "generate text". If the documents can be ocred then we need to determine what cause the issue. If the document in fact can not be ocred we can look into an alternative procedure.

Please contact your reseller so he/she can open a case on your behalf with Technical Support.

Regards

1 0
replied on May 12, 2015

Have you tried to Re-OCR the document to see if that generates the text?

1 0
replied on May 11, 2015

If the documents were OCRed, that's usually an indication that the OCR engine tried to get text but couldn't (usually because the page doesn't meet its image requirements, for ex, the resolution is under 75DPI).

0 0
replied on May 12, 2015

All pages of the shown example are 295 or 300 dpi.

0 0
replied on May 12, 2015

Yes, the documents can be OCR'd. I copied the file and Re-OCR'd it. It processed just fine. 

 

I will open a case with my VAR.

 

Thanks

0 0
You are not allowed to follow up in this post.

Sign in to reply to this post.