OCR has rare results

Discussion for the End User use uf OCR in PDF-XChange Editor and Viewer

Moderators: TrackerSupp-Daniel, Tracker Support, Vasyl-Tracker Dev Team, Sean - Tracker, Paul - Tracker Supp, Chris - Tracker Supp, Tracker Supp-Stefan, Ivan - Tracker Software

Post Reply
TGE
User
Posts: 10
Joined: Mon Feb 12, 2018 12:39 pm

OCR has rare results

Post by TGE » Thu Nov 21, 2019 3:32 pm

Hi

I've seen that the OCR in Version 8 (latest 8.0. build 334 in german) brings really rare output. To show you I have attached two documents and let the OCR run with the Option Editable Text and Images. So if you compare the pictures on Page 13 for example you see really awful things.
I know the first Option Searchable Image bings up the Document with now optical Bugs. But default is the middle Option.
I have some other documents (i can't share here) for example one with copied and scanned pages which don't brings good output from the OCR. OCR done by Ab*y or Ad*be brings a useable output. Sorry to tell.
So please check if there is something which is going wrong ....

Thanks

Tobias
Attachments
WO2018107227A1_Readable.pdf
With OCR in Editable Text and Images
(320 KiB) Downloaded 6 times
WO2018107227A1.pdf
Original
(628.9 KiB) Downloaded 5 times

User avatar
TrackerSupp-Daniel
Site Admin
Posts: 2713
Joined: Wed Jan 03, 2018 6:52 pm

Re: OCR has rare results

Post by TrackerSupp-Daniel » Mon Nov 25, 2019 9:59 pm

Hello TGE,
Thank you for the sample files here, As it is the images affected here, I am unsure how much we can do, but we can certainly take a look and see about improving this. Could I ask you to please send us a screenshot of your OCR settings before running OCR normally?
Kind regards.
Daniel McIntyre
Support Technician
Tracker Software Products (Canada) LTD

Sales: +1 (250) 324-1621
Fax: +1 (250) 324-1623

TGE
User
Posts: 10
Joined: Mon Feb 12, 2018 12:39 pm

Re: OCR has rare results

Post by TGE » Tue Nov 26, 2019 2:43 pm

Hello Daniel

The default is the Editable Text and Picture Option. But this replaces also the "picture" of the PDF so we have use the first option because its mandantory for us to hold every information. And documents as the WO***.pdf is our buisness.

And i have some scanned pages which get a good OCR from Ab*y and A*obe but the OCR from PDF-XChange isn't useable. Sorry to tell but it seems that there is something going wrong.

Regards

Tobias
Attachments
OCR_Settings.jpg

User avatar
TrackerSupp-Daniel
Site Admin
Posts: 2713
Joined: Wed Jan 03, 2018 6:52 pm

Re: OCR has rare results

Post by TrackerSupp-Daniel » Tue Nov 26, 2019 7:44 pm

Hello again TGE,

Thank you for the settings there, and I am sorry that we dont have this working quite up to par just yet. I have discussed this with the dev team and they mentioned that we are working on a new zoning function, that will allow you to correct our current "auto-zoning" function manually. I cannot say when this will be available, but it is in the works.

Currently the only way to avoid this is to do a manual OCR process, by using the snapshot tool:
  • 1. Select Snapshot tool
    2. Drag a rectangle and right click on it to choose "OCR region"
    image.png
    3. Setup OCR as desired, and then move onto the next section.
I know this is not an ideal way to do this, but it is the best I can offer at the moment.

Kind regards,
Daniel McIntyre
Support Technician
Tracker Software Products (Canada) LTD

Sales: +1 (250) 324-1621
Fax: +1 (250) 324-1623

Post Reply