OCR (ABBYY) - UB-Mannheim/Reichsanzeiger GitHub Wiki
Results from OCR with ABBYY Finereader Engine 11 for Linux
ABBYY Finereader was used by many libraries until better Open Source OCR software and models were available (for example Tesseract).
For historic prints normally a special (expensive) license is required. Here we tested ABBYY Finereader with our normal license which does a good job for layout detection, but fails to recognize the historic Fraktur script.
ABBYY Finereader is rather fast, but some special pages have processing times of more than an hour:
- https://digi.bib.uni-mannheim.de/periodika/fileadmin/data/DeutReunP_856399094_19231120/max/856399094_1923_265_27.jpg (1:45 h)
- https://digi.bib.uni-mannheim.de/periodika/fileadmin/data/DeutReunP_856399094_19231206/max/856399094_1923_278_39.jpg (1:36 h)