So, in most cases, pages that do not meet the First Page Identification conditions for any document classes will be appended as additional pages to the current document--the document that was created when a page satisfied a document class's First Page Identification conditions.

If the document is always 2 pages, the easiest way to break the document up is to set the Last Page Identification page limit to 2 and remove any Last Page Identification conditions (and maybe remove First Page Identification conditions, depending on your situation).

If the situation isn't that simple, you'll need to find a way to make your First Page Identification condition specific enough that the second page is not identified as a first page. 

Based on your post it sounds like you also might have Last Page Identification conditions configured? If so, you may need to adjust those or remove them if they are identifying the first page as the last page.

Another thing that can affect the way pages are assigned to documents is the "Keep each file as a separate document" options in the scan source options.

Hi Christian, 

Can you take screenshot of your First Page Identification or Last Page Identification and upload them (You can erase any sensitive info and only show the configuration).

And if you are using First Page Identification, have you tried to add a OmniPage Zone OCR process?

In your First Page Identification, add zone to where text "admin" is. And set condition to "Identification (Zone 1) is empty" so that it will identify page 1 but not page 2.