Hi, What type of filters do you use? Thank you. Regards,
Georg Oberth [email protected] www.godata.at 0043-676-692-6070 0043-3127-88121 -----Ursprüngliche Nachricht----- Von: [email protected] [mailto:[email protected]] Im Auftrag von Svetlin Nakov Gesendet: Saturday, November 14, 2009 5:51 PM An: [email protected] Betreff: RE: Text inserted by camara Hello, you need to apply some special filters to remove the background and the text outline and to keep the text only (the white color on the picture). After that you need to train Tesseract for exactly this font (same font name, same size, same color, same character spacing, etc. It is important to compile a comprehensive dictionary (if possible). This process will improve the results but 100% accuracy is unlikely to be achieved (with Tesseract). Note that Tesseract is designed for 300 DPI images with font size of 8pt - 64pt and when the image has less DPI (even when the quality is perfect) or the font is relatively smaller it does not perform well (* this conclusion is unofficial, just my personal experience). Svetlin Nakov Managing Partner Consulting and Information Technology Agency (CITA) http://www.citagency.eu Author of the book for beginner Java develoepers: http://www.introprogramming.info -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Lucas U6 Sent: Saturday, November 14, 2009 1:57 PM To: tesseract-ocr Subject: Text inserted by camara Hi, I am running tests on images with text inserted by the camera. in some cases it works perfect but most do not get good results. How can I improve reading? Here an example and read the ocr Image http://i33.tinypic.com/2djuj5.jpg Debug capture http://i33.tinypic.com/2rh9yk5.jpg Greetings. Lucas ____________________________________________________________________________ __________________ Hola, Estoy ejecutando pruebas sobre imagenes con texto insertado por la camara. en algunos casos funciona perfecto pero en la mayoria no obtengo buenos resultados. Como puedo mejorar la lectura? aqui un ejemplo y que leyo el ocr Imagen http://i33.tinypic.com/2djuj5.jpg Debug captura http://i33.tinypic.com/2rh9yk5.jpg --~--~---------~--~----~------------~-------~--~----~ 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 -~----------~----~----~----~------~----~------~--~---
