女书联机手写识别样本采集系统的设计与实现

Design and Implementation of an Sample Collection System for Online Handwritten NǜShu Recognition

女书联机手写识别是一种实时识别女书的方式，用户一边写，机器一边识别处理。使用者只要在书写板上按常规方法书写，就可以把女书输入计算机进行辨识，方便用户使用。为提高识别率，识别前需采集不同人书写的女书样本进行特征提取，从而建立女书特征库字典。女书联机手写识别样本采集系统将采集到的女书字符信息进行预处理后保存到数据文件中．并实现对数据文件的读取以及文字的动态还原。

Online Handwritten NǜShu Recognition is a real-time recognition method, NǜShu characters can be recognized when users are writing on a handwritten board. To improve the recognition rate, we extract features from collected NǜShu samples written by different users, then establish the NuShu character library. This sample collection system saves collected NuShu character information after preprocessing into data files, and reads them back to regenerate written characters dynamically.