Automatic extraction of correlation-entropy features for text document analysis directly in run-length compressed domain.
Mohammed JavedP. NagabhushanB. B. ChaudhuriPublished in: ICDAR (2015)
Keyphrases
- document analysis
- automatic extraction
- run length
- compressed domain
- document images
- video analysis
- image analysis
- character recognition
- gray level
- writer identification
- text analysis
- co occurrence
- feature extraction
- image compression
- image features
- feature vectors
- pattern recognition
- information retrieval
- machine learning
- text mining
- inter frame
- feature space
- multiscale
- dct domain