Adapting the Tesseract Open Source OCR Engine for Multilingual OCR

   Abstract