tesseract-ocr / langdata

Source training data for Tesseract for lots of languages
Apache License 2.0
826 stars 886 forks source link

About Uyghur Language recognition #149

Open rustam opened 4 years ago

rustam commented 4 years ago

The accuracy of Uyghur(Uighur) Language recognition increased dramatically, but when it comes to recognize fonts besides the one which is being used for training, the accuracy was dropped and even can't recognize at all. strongly recommend to use the following fonts provided by UKIJ ( http://www.ukij.org) which is commonly used by people .

UKIJ Tuz : http://www.ukij.org/fonts/fonts/UKIJTuz.ttf UKIJ Tuz(Bold) : http://www.ukij.org/fonts/fonts/UKIJTuzBold.ttf UKIJ Nasq : http://www.ukij.org/fonts/fonts/UKIJNsq.ttf UKIJ Nasq(Bold) : http://www.ukij.org/fonts/fonts/UKIJNsqb.ttf UKIJ Basma : http://www.ukij.org/fonts/fonts/UKIJBasma.ttf UKIJ Zilwa : http://www.ukij.org/fonts/fonts/UKIJZilwa.ttf UKIJ Esliye : http://www.ukij.org/fonts/fonts/UKIJEs.ttf UKIJ Esliye(Bold) : http://www.ukij.org/fonts/fonts/UKIJEsBold.ttf UKIJ Tuz Basma : http://www.ukij.org/fonts/fonts/UKIJTuzB.ttf UKIJ Tuz Basma(Bold) : http://www.ukij.org/fonts/fonts/UKIJTuzBB.ttf UKIJ Tuz Kitab : http://www.ukij.org/fonts/fonts/UKIJTuzK.ttf UKIJ Tuz Kitab(bold) : http://www.ukij.org/fonts/fonts/UKIJTuzKB.ttf UKIJ Tuz Gezit : http://www.ukij.org/fonts/fonts/UKIJTuzG.ttf UKIJ Tuz Gezit(bold) : http://www.ukij.org/fonts/fonts/UKIJTuzGB.ttf UKIJ Tuz Qara : http://www.ukij.org/fonts/fonts/UKIJTuzQ.ttf UKIJ Tuz Qara(Bold) : http://www.ukij.org/fonts/fonts/UKIJTuzQB.ttf UKIJ Tuz Tor : http://www.ukij.org/fonts/fonts/UKIJTzTr.ttf UKIJ Tuz Tor(Bold) : http://www.ukij.org/fonts/fonts/UKIJTzTrBold.ttf UKIJ Qara : http://www.ukij.org/fonts/fonts/UKIJQara.ttf UKIJ Qara(Bold) : http://www.ukij.org/fonts/fonts/UKIJQara-b.ttf