Examining a multi-layered approach for classification of Optical Character Recognition (OCR) quality without Ground Truth
We are examing a multi-layered approach for the classification of digitized texts, without the presence of Ground Truth (manually corrected) texts.
Verfasser: | |
---|---|
Dokumenttyp: | conferencePaper |
Erscheinungsdatum: | 2021 |
Schlagwörter: | OCR / Dutch / Text analysis / Quality |
Sprache: | unknown |
Permalink: | https://search.fid-benelux.de/Record/base-26689336 |
Datenquelle: | BASE; Originalkatalog |
Powered By: | BASE |
Link(s) : | https://zenodo.org/record/4896225 |