/interactive-ocr

Implementation of a couple of heuristics that estimate OCR quality without reliance on ground truth data, focusing on historical documents written in English.

Primary LanguagePython

interactive-ocr

cleaning this up is on my postgrad todo list