Quote:
Originally Posted by forcheville
I'm currently scanning a paper book that I purchased so that I can read it on my portable device. It's around 600p and I've been doing about 100p/week and correcting as I go.
It's a lot more work than I expected when I started as the print quality in this particular volume (a recent edition of a European classic first published in 1902 and still in print) is appalling. It is printed on coarse paper such that the ink often seems to bleed out beyond the letter outline. There are other problems as well, and they all make for a high density of errors even with the best OCR software I can find.
|
You may already have done so, but have you tried training the software recognition? I've found that that helps a good deal with old books.