The offline recognition of Chinese Characters is a very important research field in OCR. In this paper, the authors propose an effective stroke extraction model for Chinese Character based on Stroke Segmentation and Combination. Comparing with the existed models, the experiments proved that it could improve the precision for the stroke extraction and reduce the computation complexity considerably.
Published in:
Machine Learning and Cybernetics, 2003 International Conference on
(Volume:5
)
Date of Conference: 2-5 Nov. 2003