摘要
该文提出了一种基于DP算法和隐马尔可夫模型的汉字手写体识别方法,通过提取整字特征和笔划特征来描述汉字特征信息,识别时,采用DP匹配算法,使得字库样本与待识样本的码列匹配关系是最优的,当出现连笔系统拒识时采用隐马尔可夫整字分类器,从而提高整体的识别效率。
A plan about the handwriting input method base on dynamic programming（DP） and hidden Markov model is introduced in this paper. The whole features and the strokes features are extracted to describe the Chinese character. During the recognition, we use dynamic programming to make the matching of font stylebook and pending stylebook optimization, if the words written by consecutive cause the exclusion to appear unexpectedly. HMM base on whole classifier is used to improve the whole recognition rate.
出处
《杭州电子科技大学学报（自然科学版）》
2008年第5期120-123,共4页
Journal of Hangzhou Dianzi University：Natural Sciences
基金
国家自然科学基金资助项目(60671037)
浙江省重大科技攻关资助项目(C11200)
宁波市工业攻关计划资助项目(B10051)
