rigaudon
rigaudon copied to clipboard
Polytonic Greek OCR engine derived from Gamera and based on the work of Dalitz and Brandt
Id is based on pageId, classifier, gamera_code_version, linesegmentation_approach, then we also give character, bbox and confidence
Somehow we are outputting combined unicode letters. In this case, we should provide greek.small.letter.alpha.with.oxia (x1F71) not greek.small.letter.alpha.with.tonos (x03AC).
Unicode output is decomposed. Federico's analyser requires composed unicode, so we need to provide composed output.