Highly accurate recognition of printed Korean characters through an improved two-stage classification method.
Jin-Soo LeeOh Jun KwonSung Yang BangPublished in: Pattern Recognit. (1999)
Keyphrases
- highly accurate
- classification method
- handwritten characters
- optical character recognition
- knn
- recognition rate
- k nearest neighbor
- printed characters
- character segmentation
- capable of producing
- support vector machine
- text classification
- high quality
- support vector machine svm
- character recognition
- classification scheme
- automatic recognition
- license plate
- high accuracy
- pattern recognition
- feature extraction
- handwritten words
- accurate models
- data sets
- databases
- machine learning
- information retrieval
- associative classification
- handwriting recognition
- database
- reinforcement learning
- test sample
- printed documents
- information extraction
- small number
- nearest neighbor