Arabic traineddata for 3.0x uses cube engine. Training process for that was
never shared. Now the cube engine has been removed for lstm 4.0, which is
still in alpha stage.
There is 4.0alpha traineddata for Arabic and you can train for it , but
accuracy is not great. Ray is doing another training
Hello All,
I want to ask about the issue that I faced after making training for
tesseract for *Arabic*, I have Ubuntu & Tesseract-ocr 3.04
*Steps*:
1. $ convert ara.arial.exp1.jpg ara.arial.exp1.tif
2. $ tesseract ara.arial.exp1.tif ara.arial.exp1 -l ara batch.nochop
makebox
3.
2 matches
Mail list logo