تيسراكت (Tesseract) لاستخراج النص من الصور

لم ينشر أحد أي قياس لهذا.

• العربية الفصحى

  Tesseract has published Arabic-language trained data since version 3 (https://github.com/tesseract-ocr/tessdata), and community reports describe known problems specifically with Arabic and other right-to-left scripts, but no official per-language accuracy figure is published and no independent benchmark was run for this entry. Untested: expect to need to review and correct its output on scanned Arabic text, especially handwriting or low-quality scans.

• اللهجة العراقية

  OCR reads printed or handwritten script rather than spoken dialect, so Iraqi-dialect text written in standard Arabic script is covered by the same trained data as MSA, at the same unverified accuracy. Untested.

• الكردية السورانية

  Community-contributed Kurdish trained data exists for Tesseract but is not part of the officially maintained language packs checked for this entry, and no benchmark was run. Untested.