随着我国计算机技术的发展，少数民族信息处理已经逐渐成熟起来，少数民族文字识别研究已经成为一个热点。本文提出一种基于小波包特征与径向基网络的少数民族文字种类识别方法，该方法采用小波包能量和小波包能量比例分布的特征描述，利用径向基函数神经网络对少数民族文种进行分类识别。通过构建六种常用的少数民族文字及汉字、英文共八种文字的样本库，采用本文的方法对样本库进行了训练和测试。实验结果显示，本文的方法对于少数民族文种识别的平均精度好于小波特征及传统的分类方法。

With the fast development of computer technology, Chinese minority information processing has become mature gradually, and the research on Chinese minority information processing focuses on the Chinese minority optical character recognition. The method of identification of the kinds of Chinese minority scripts based on wavelet packet analysis and Radial Basis Function Network (RBFN) is presented which adopts the feature of wavelet packet energy and wavelet packet energy distribution proportion, and constructs multivariate classification in the radial basis function Network. By building a data set which contains 6 kinds of common Chinese minority scripts, 8 kinds of Chinese and English in total, we train and test the dataset by means of the method in this paper. Obviously, the result shows that the method outperforms the traditional classification and wavelet feature.