Are you referring to how the second number "1" appears different from the first number "1?"
It appears to me that one was recognized by the scanning software and the other was not.
Logically if something like that were to happen legitimately, it would mean that the trailing '1' had registered as part of a background layer which was treated as an image, and that the other digits had been scanned as text and added to a text layer.
Now, again logically, you would expect that to happen if the user had scanned the document as editable text rather than as an image. But in such a case the document would at least be searchable, and this one isn't. You can't even search for other parts of the digit string, say, "106". Try it; try searching on any text string you see in the document.