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

Question

Question

Different Results from OmniPage OCR and OmniPage Zone OCR

Quick Fields
Updated September 18, 2015
asked on September 18, 2015

I have a form that I am running through Quick Fields that identifies some text on a page and then pattern matches the text  out to find the value I'm looking for.  In doing so, I'm getting different results using different OCR types.  As you can see in the screenshot below, the Text (OmniPage OCR) reads the document correctly, but then at the bottom you can see that the Zone OCR (Shipment ID) activity does not identify this correctly.  Here are the differences I can see from looking into this.  

1.  Zone OCR is set to have LETTERS as the character preference - I would think this would give me BETTER results, but it is not!  

2.  OmniPage OCR is set to decolumnize text - I do not see this option in the Zone OCR activity.  

I'm curious if anyone else is seeing this, and if so, what is the fix?  You cannot use OmniPage OCR as a token to Pattern Match against, so I am stuck using the undesired text.  Any help is much appreciated!  

Thanks,

Nate

0 0

Replies

replied on September 18, 2015

You can use the OCRed text in Pattern Matching.

The OCR engine does make some guesses based on the amount of text in the region. Try using a Smooth process as a local image enhancement to Zone OCR to clean up the characters in the region a bit and see if it make a difference. Or try setting Zone OCR's optimization settings to "accuracy".

1 0
You are not allowed to follow up in this post.

Sign in to reply to this post.

Details
Related Posts
Loading ...