基于子笔划的手写体汉字识别

Handwritten Chinese Character Recognition Based on Sub-stroke

该文给出一种汉字子笔划的提取方法 ,较好地解决了因子笔划相交 ,使提取的同类字符子笔划不稳定问题。引入了字符的固定部件划分方法及特征矩阵的概念。字符部件中包含各类子笔划的位置、长度等信息 ,特征矩阵是这些信息的集中体现。设计了一种新的遗传算法 ,由未知字符一个部件的一类子笔划表示的字符与某类模型的相似度 ,既由与其它 3类子笔划相应的相似度调节 ,又由与周围部件同类子笔划相应的相似度调节。实验表明 。

A new approach to extract the sub stroke of handwritten Chinese character is proposed, which solves the non stability problem of sub stroke extracted from characters of the same class. The method of separating the fixed components of character and the concept of eigenmatrix of character are introduced. The information in character components includes the position and the length of different classes of sub strokes, all of which were embodied in the eigenmatrix. A new genetic algorithm is developed and used in character recognition. The similarity between the characters represented by the sub stroke class of the component of an unknown character and the model of a certain class is regulated not only by the similarity with the other three classes of sub strokes respectively but also by the corresponding similarity with the sub strokes of the same class of the peripheral components. Experiment shows that the new approach is effective for handwritten Chinese character recognition.