Hello Antonio I have made my self a bookreading device, so I am scanning a lot of books. I am reading mainly names of persons and dates.
I am using Omnipage which seems to have an excellent OCR engine, but the result gets destroyed by its (intelligence). A simple 2 column text gets garbled by a much to smart zoning algoritme, so manually correction are needed. You can't get the text as is, Omnipage adds and removes linebreaks spaces and so on. Omnipage don't read a characterset like iso-8859-15, but a language so you have to enable the languages holding the needed characters, after which the spell checkers destroys the result. A characters close to numbers gets changed to numbers, and a number close to characters gets changed to characters. Omnipage don't have a commandline option and the graphical userinterface are at the least a waste of time. So I would like to use OCRAD, but there don't seem to be a Windows binary ? I probably could build it myself, but I am from the hardware end, so it may take forever. I have been using UNIX/LINUX/GNU tools for many years mostly under windows. Regards John Nielsen _______________________________________________ Bug-ocrad mailing list [email protected] http://lists.gnu.org/mailman/listinfo/bug-ocrad
