Omnifont Persian OCR System Using Primitives.
Azarakhsh KeipourMohammad EshghiSina Mohammadzadeh GhadikolaeiNegin MohammadiShahab EnsafiPublished in: CoRR (2022)
Keyphrases
- optical character recognition
- document images
- post processing
- recognition errors
- character recognition
- error correction
- preprocessing
- text recognition
- scanned documents
- document processing
- knowledge base
- text classification
- machine learning
- document image retrieval
- printed documents
- text retrieval
- low level
- visual patterns
- end to end
- building blocks
- query expansion
- high level
- data sets