[bksvol-discuss] Any Difference in OpenBook's OCR Engines?
- From: "Evan Reese" <mentat1@xxxxxxxxxxxxxx>
- To: <bksvol-discuss@xxxxxxxxxxxxx>
- Date: Thu, 9 Feb 2006 10:15:18 -0800
I was just curious if any OpenBook users had noticed any difference in the
three OCR engines it uses. I tried scanning a page or two using all three and
I couldn't really tell any difference. But I've only scanned three books so
far, all using Fine Reader as the default. But those with more experience
might be more aware of differences that might not show up in a page or two.
Even a very slight difference that's hard to detect will add up to quite a few
errors after three- or four-hundred pages.
If I get any responses, and if I see a consensus brewing that either omnipage
or Recognita is better, I will change to one of those. I get pretty good
results with Fine Reader. But it could always be better, and I'll change in a
moment if more experienced people think one of the others is better. If they
don't agree, of course, then I'll stick with Fine Reader.
Also, another reason I ask is that the book I'm reading through is of very high
quality. Once the preliminaries are over, the text is extremely good. It
might be better than my OpenBook could do. Of the two hardcovers I have
scanned so far, they came out pretty well, but this might - might - be better.
I don't know what system he is using, but it is very good. It still misses
some things, of course. It's not perfect, especially those
beginning-of-chapter first words that no scanner seems to be able to get right.
But there are quite a few pages here with nothing whatever wrong on them. I
admit, I'm a little jealous. I get perfect pages, too, but I don't think I get
as many as this guy gets. Sure, not all books are created equal, so maybe he
got an extremely excellent font in this book. But maybe I'm not using the best
OCR engine, or the best settings I could be. I've just been using automatic
contrast on the books I've done so far.
Thanks for any feedback.
Other related posts:
