手写体汉字特征的抽取研究

On Feature Abstraction for Handwritten Chinese Characters

本文采用结构方法研究汉字手写体识别。通过利用全新的防噪声技术,有效地抽取了规范手写体汉字的自然笔划并自动形成结构特征编码。最后,我们对基于人工神经网络的手写体汉字识别进行了初步探讨。

In this paper a structural method is used for the recognition of hand- written Chinese characters.By using a novel noise-proof technique,we can effcetively ab- stract the natrual strokes of regular handwritten Chinese characters and realize the structural feature encoding.Finally,We discuss the ANN-based recognition of hand written Chinese characters.