Document image segmentation using discriminative learning over connected components.
Syed Saqib BukhariMayce Ibrahim Ali Al AzawiFaisal ShafaitThomas M. BreuelPublished in: Document Analysis Systems (2010)
Keyphrases
- connected components
- discriminative learning
- image segmentation
- level set
- watershed transformation
- binary images
- generative model
- weakly supervised
- document images
- text lines
- active contours
- energy function
- document classification
- object categories
- pairwise constraints
- shape prior
- graph cuts
- document collections
- image set
- multiscale
- document clustering
- maximum margin
- markov random field
- image processing
- segmentation method
- data sets
- highly efficient
- gray scale
- segmentation algorithm
- structured prediction
- computer vision
- maximum likelihood
- data structure
- active learning