Extraction of text-related features for condensing image documents.
Dan S. BloombergFrancine ChenPublished in: Document Recognition (1996)
Keyphrases
- image features
- automatically extracted
- test images
- extracted features
- low level
- scanned documents
- input image
- free text
- information retrieval
- text lines
- spatial information
- semantic information
- image representation
- keywords
- feature extraction
- text information
- multiscale
- feature vectors
- document analysis
- textual features
- image content
- web images
- image retrieval
- text documents
- printed documents
- web documents
- information extraction
- digital documents
- feature space
- latent semantic analysis
- textual content
- text regions
- keypoints
- co occurrence
- image classification
- related documents
- image segmentation
- linguistic information
- semantically related
- text collections
- text retrieval
- image regions
- feature points
- textual descriptions
- query expansion
- feature set
- text extraction
- printed text