脱机手写汉字中叉点精细化的改进算法

Improved Algorithm of Fork Point Optimization for Offline Handwritten Chinese Character

提出了一种新的脱机手写汉字中叉点精细化的改进算法。它主要是在提取了手写汉字骨架之后,用Rutoviz相交数得到特征点。在此基础上进行叉点精细化的算法。该改进算法以最大圆规则为基础,结合实验中发现的问题,增加了新的规则,使叉点结合的错误率几乎为零。从而进一步提高了脱机手写汉字识别的正确率。

This paper proposes a new optimal algorithm of fork points in off-line handwritten Chinese character. The algorithm is processed after getting feature points by using Retoviz＇ definition of crossing number. This algorithm improves maximum circle criterion, reduces further error rate of grouping fork point. Hence, improve the correct rate of off-line handwritten Chinese character recognition.