VML-MOC: Segmenting a multiply oriented and curved handwritten text lines dataset.
Berat Kurar BarakatRafi CohenIrina RabaevJihad El-SanaPublished in: CoRR (2021)
Keyphrases
- text lines
- text line segmentation
- document images
- handwritten documents
- handwriting recognition
- optical character recognition
- text detection
- character recognition
- connected components
- document image analysis
- printed documents
- line extraction
- complex background
- hough transform
- face recognition
- feature set
- structural features