tesseract osd retrain for other undected script which is available in script Dir

[omesh-sharma]

Hey

---

i am Using Tesseract OCR for the text extraction form the image :

---

---

I need your valuable suggestion for the below mentioned points.

---

• How can i Retrain osd.traindata file for adding Ethiopic and other scripts , because current osd.traindata file unable to detect few scripts name eg:(ethiopic , gujarati, gurmukhi) but script files for them are available in script directory.

---

---

• which is more accurate for text extraction [LANGUAGE TRAIN DATA FILES] or [SCRIPT TRAIN DATA FILES]

---

---

• Does it make any difference to use the script for text extraction instead of language.traindata in term of text extraction accuracy.

---

---

Please Share your valuable comments and suggestions for above mentioned list as per your experience with tesseract.
It'll be very helpful for my final year project.

---

Contact: sharmaomesh0@gmail.com .

Thanks and regards
Omesh sharma