It is possible, and there are broken bits of code that support that kind of training, but it hasn't been used for years and no longer works, so it would take quite a lot of effort to get it working.Ray.
On Thu, Nov 20, 2008 at 4:29 AM, Philipp Lenssen <[EMAIL PROTECTED]>wrote: > > Hi! I read through (http://code.google.com/p/tesseract-ocr/wiki/ > TrainingTesseract) but wanted to see if there's an easier option than > creating specific bounding boxes for each letter (which is what I > understand the tutorial says one needs to do?). Is there any option > where one would simply point to a TIF and TXT file, the TXT file > containing the correct text, and thus train Tesseract accordingly? > > For instance, I'm currently getting a result like this one on an > image: > ------------ > Aprll 15 1953 > Foober > ------------ > > So I would like to change the text to > ------------ > April 15 1953 > Foobar > ------------ > ... for training purposes (guessing that Tesseract could take a try at > figuring out the bounding boxes itself as it did for the first > incorrect run?). > > Thanks! > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/tesseract-ocr?hl=en -~----------~----~----~----~------~----~------~--~---
