Historical Document Text Binarization using Atrous Convolution and Multi-Scale Feature Decoder.
Hanif RasyidiSalman KhanPublished in: DICTA (2019)
Keyphrases
- document images
- multiscale
- historical documents
- document processing
- document analysis
- text documents
- web documents
- keywords
- information retrieval
- digital documents
- printed documents
- document content
- text content
- text lines
- text mining
- textual content
- historical manuscripts
- scanned documents
- image processing
- text collections
- semantic information
- scientific documents
- document structure
- text retrieval
- text clustering
- document set
- text summarization
- image features
- document collections
- text categorization
- feature vectors
- document corpus
- feature set
- information retrieval systems
- handwritten documents
- text corpus
- wordnet
- pdf files
- handwritten text
- scientific papers
- wavelet transform
- edge detection
- scale space
- free text
- document level
- image representation
- handwriting recognition
- sentence level
- latent semantic analysis
- text recognition
- image segmentation
- document clustering
- text classifiers
- automatic text summarization
- optical character recognition
- multimedia documents
- character recognition
- document retrieval
- co occurrence
- distributed video coding