Word and character segmentation is performed directly on run-length encoded handwritten document images, with 90.29% word and 86.23% character accuracy on ICDAR13 English handwritten data.
The word segmentation approach is based on the inter-word gaps whereas the character segmentation is based on the intra -word spacing
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Word and character segmentation directly in run-length compressed handwritten document images
Word and character segmentation is performed directly on run-length encoded handwritten document images, with 90.29% word and 86.23% character accuracy on ICDAR13 English handwritten data.