Recognition of Chinese Characters by MomentFeature

To extract a set of suitable features from all Chinese characters for the purpose of classiication is essential in the design of an Optical Character Recognition (OCR) system for printed Chinese characters. Moment descriptors have been developed as features in pattern recognition since the moment method was rst introduced. The moment method is diierent from other feature extraction methods used in Chinese character recognition, in that, moments can be carefully selected such that the extracted features are invariant under translation, scale change, and rotation. The moment method captures global properties of characters rather than local ones. In this paper, new moment features for Chinese character recognition are proposed which provide signiicant improvements in terms of Chinese character recognition, especially for those characters that are very close in shapes.