Re: Document Scanner


 

I will second Gene's observation that I often see worse OCR results when I use a "high accuracy" setting than I do with a medium or low.  And the reason for this, I believe, is that "high accuracy" settings involve rejection of anything "questionable" while lower accuracy settings accept it.  When you're talking about printed material being scanned in the first place, allowing for "low accuracy" will get you maximal recognition of what's there, and what has to be at least reasonably clear to begin with.

I've seen OCR work what I almost feel are miracles on some awful "10th generation" scans of original documents that are not all that easy to read from the start.

I also haven't seen all that much difference in accuracy when the original material being scanned is very clean to begin with.  But, if you want the ability to mask off parts of pages (e.g. title line and page numbers at the top or footnotes at the bottom) you'll have to use specialty software if you're using a document feeder as it's impossible to create a mask under that circumstance.

--
Brian  Windows 10 Home, 64-Bit, Version 1703, Build 15063  (dot level on request - it changes too often to keep in signature)

     The opposite of a correct statement is a false statement.  But the opposite of a profound truth may well be another profound truth.

            Niels Bohr

 

 

Join nvda@nvda.groups.io to automatically receive all group messages.