OCR is not accurate for Persian ( fas ) language
47 views
Skip to first unread message
MohamadAli Ganji
Jul 13, 2016, 6:59:46 AM7/13/16
to tesseract-ocr
Hello,
I just find this good OCR and i used it for my language (Persian) it`s coded as `fas`, but the result is terrible.
I checked `tessdata` in github repo but it`s seems it`s incomplete by comparing to other languages.
Here is I am asking for help please :
1.Why the OCR does not work well with this language ? and how may improve it ?
2.if its related to `tessdata` or need something like to be trained, how may i do it ?
Thank you very much.
I just find this good OCR and i used it for my language (Persian) it`s coded as `fas`, but the result is terrible.
I checked `tessdata` in github repo but it`s seems it`s incomplete by comparing to other languages.
Here is I am asking for help please :
1.Why the OCR does not work well with this language ? and how may improve it ?
2.if its related to `tessdata` or need something like to be trained, how may i do it ?
Thank you very much.
Reply all
Reply to author
Forward
0 new messages