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

Question

Question

Redefining page number in QF with token or condition

Quick Fields Ideas
Updated July 1, 2014
asked on June 25, 2014

I have seen a few questions/responses around this topic, but not quite answering this:

 

Details

We import documents (not electronic) from our repository into quickfields.  Each document is distinct with various numbers of pages.  I need to OMR from one of those pages (page 4), however, it is not always the 4th page after staff rearranges the order.

 

To make this as efficient as possible, is there a way to identify that page with ZoneOCR in preclassification, assign it a token if it meets the criteria for page 4, and then somehow have Quickfields recognize that page as page 4 instead of whatever page it was?  I would like to do this to ensure OMR just acts on the correct page regardless of its original page position.

 

To me, tasks like Form Identification or ZoneOCR should be able to recognize that page in preclass, store it as a match, and then be able to pass it off to OMR without requiring a single page number, or all pages.

 

The capture process for any organization is not typically neat, or sequential.  So the only reason I could see for only allowing precise page numbers for OMR is to upsell the Conditional feature for quickfields.

 

Any thoughts on how to accomplish this without buying every add-on?  It may be right in front of me and just need tips from someone who has encountered this type of issue before.

 

Thanks in advance!

 

0 0

Answer

SELECTED ANSWER
replied on June 26, 2014

I assume we're doing this with a single document class (based on your QF configuration listed in our database). It would be easier with 2, but you can get pretty close with one.

 

Set Capture Engine to keep the document structure. Set the document class identification processes so either the first page or the page you're after succeeds. Leave Last Page Identification on its default of "when first page is identified and a new document is created". Then under Page Processing, set OCR to page 1.This will result in documents from LF being broken into 2 docs: page 1 to page before the one you want and page you want to last page. The first page of each one will be OCRed. If you're not doing any other processing, these 2 documents will both have the same name.

 

Set the document class to merge documents with the same name when they're stored. This will recombine the 2 documents when sent to LF.
 

0 0

Replies

replied on June 25, 2014

This is not currently possible without a Conditional Process.

 

Quick Fields does not run processes in document classes until the document is identified as being part of the document class. This is by design in order to achieve better performance (as running all possible processes on all possible pages would result in a lot of wasted work.

 

So while you can read data from the page with Zone OCR as part of Pre-Classification, the Conditional process is what allows you to tie results from this process to the following ones when a specific page number is not possible.

0 0
replied on June 26, 2014

Thank you for the reply.  So if I am only trying to pluck out one page to OCR from, let's say an 11 page scanned document, could I set both the first and last page identifiers to look for the page I want to OCR?  Then perform OCR on that page to store along with the OCR tokens?  The documents are being read from those already stored in the repository.

 

Just trying to find the easiest way to pick a needle out of the haystack, see that it is the right needle, and then capture details about the needle and move to the next haystack.  Sorry for the bad analogy :)

0 0
SELECTED ANSWER
replied on June 26, 2014

I assume we're doing this with a single document class (based on your QF configuration listed in our database). It would be easier with 2, but you can get pretty close with one.

 

Set Capture Engine to keep the document structure. Set the document class identification processes so either the first page or the page you're after succeeds. Leave Last Page Identification on its default of "when first page is identified and a new document is created". Then under Page Processing, set OCR to page 1.This will result in documents from LF being broken into 2 docs: page 1 to page before the one you want and page you want to last page. The first page of each one will be OCRed. If you're not doing any other processing, these 2 documents will both have the same name.

 

Set the document class to merge documents with the same name when they're stored. This will recombine the 2 documents when sent to LF.
 

0 0
replied on July 1, 2014

Thank you Miruna.  I essentially did that, just with defining the start and end conditions to match the page I was looking to OMR.  I then created a new document with OMR in the name with just that page and the recorded token values.  For now it works, but in the future, we are going to require page numbers on the document to make identification easier.

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

Sign in to reply to this post.

Details
Related Posts
Loading ...