Robust feature extraction method based on run-length compensation for degraded character recognition.
Minoru MoriMinako SawakiNorihiro HagitaHiroshi MuraseNaoki MukawaPublished in: Systems and Computers in Japan (2004)
Keyphrases
- character recognition
- run length
- machine vision
- hand written
- handwritten characters
- gray level
- computer vision
- optical character recognition
- chinese characters
- sample size
- image features
- feature vectors
- character segmentation
- color images
- data hiding
- multiscale
- handwriting recognition
- image sequences
- feature selection
- machine learning
- ocr systems
- printed characters