Version 5.2.0-1 of packages libtesseract-ocr_5 tesseract-ocr tesseract-ocr-devel tesseract-training-util are available in the Cygwin distribution: Other language specific data are available upstream https://github.com/tesseract-ocr/tessdata/ while training data for building new language data are in https://github.com/tesseract-ocr/langdata CHANGES Upstream last release https://github.com/tesseract-ocr/tesseract/releases DESCRIPTION Tesseract is probably the most accurate open source OCR engine available. Combined with the Leptonica Image Processing Library it can read a wide variety of image formats and convert them to text in over 60 languages. It was one of the top 3 engines in the 1995 UNLV Accuracy test. Improved extensively by Google. It is released under the Apache License 2.0. HOMEPAGE https://github.com/tesseract-ocr/ Marco Atzeri If you have questions or comments, please send them to the cygwin mailing list at: cygwin (at) cygwin (dot) com .