Thanks for spotting this! I understand why it makes this error, but it will take some thought to fix it properly! It is using a sort by x-position to re-order the boxes for RTL language training, but that doesn't work in the case of heavily kerned characters like ل in your example. It needs to simply reverse the RTL characters, but has to avoid messing up the order of the common script, which is why I was using a sort to begin with. https://github.com/tesseract-ocr/tesseract/blob/master/training/boxchar.cpp#L202
On Thursday, March 9, 2017 at 5:11:49 AM UTC-8, El Fakir Zakaria wrote: > > I noticed that tesseract4 reads الأ as األ which is pretty close, because > we need to switch the position of the last 2 letters to have ا ل أ, this > happens with similar word forms too like لا reads as ال and should be ل ا, > and i wish to correct it. > can someone show me how to fix this, or maybe update arabic data. > thank you for your time. > -- You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To unsubscribe from this group and stop receiving emails from it, send an email to tesseract-ocr+unsubscr...@googlegroups.com. To post to this group, send email to tesseract-ocr@googlegroups.com. Visit this group at https://groups.google.com/group/tesseract-ocr. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/d993e1d4-1978-40f8-9917-331613925457%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.