摘要
本文提出了一种通用的基于部分空间方法的相似汉字识别算法,该算法无须事先确定相似字组,也不必人工选择各个相似字组的部分空间,能够自动决定待识别字是否需要进入相似字识别过程,以及怎样选择部分空间。实验结果证明了本算法的有效性。
An Algorithm of recognition of similar Chinese characters based on part space was proposed in this paper. Using this algorithm, the sets of similar characters and the part space for each set are not need to be pre-decided, whether an input character should enter the procedure of the recognition of similar characters and how to select the part space can be decided automatically. The validity of the algorithm was verified by our experiments.
出处
《中文信息学报》
CSCD
北大核心
2002年第5期44-48,共5页
Journal of Chinese Information Processing
基金
国家863计划(2001AA114080)
