Tetrak OCR
GitHub ↗
Reference

Reference

The detail behind the research: the corpus itself, what each engine does and does not do, how routing decides, and the command line.