王海燕, 王红军, 徐小力. 基于支持向量机的纳西东巴象形文字符识别*[J]. 云南大学学报(自然科学版), 2016, 38(5): 730-736. doi: 10.7540/j.ynu.20150757
引用本文: 王海燕, 王红军, 徐小力. 基于支持向量机的纳西东巴象形文字符识别*[J]. 云南大学学报(自然科学版), 2016, 38(5): 730-736. doi: 10.7540/j.ynu.20150757
WANG Hai-yan, WANG Hong-jun, XU Xiao-li. Recognition of Naxi Dongba pictographs based on support vector machine[J]. Journal of Yunnan University: Natural Sciences Edition, 2016, 38(5): 730-736. DOI: 10.7540/j.ynu.20150757
Citation: WANG Hai-yan, WANG Hong-jun, XU Xiao-li. Recognition of Naxi Dongba pictographs based on support vector machine[J]. Journal of Yunnan University: Natural Sciences Edition, 2016, 38(5): 730-736. DOI: 10.7540/j.ynu.20150757

基于支持向量机的纳西东巴象形文字符识别*

Recognition of Naxi Dongba pictographs based on support vector machine

  • 摘要: 字符识别是模式识别领域非常重要的一个应用领域.纳西东巴象形文字是世界上唯一活着的象形文字,在识别领域的研究较少.主要提取纳西东巴象形文字的粗网格特征和拓扑特征的组合,使用支持向量机方法,对纳西东巴文字进行识别.通过实验证明,采用组合特征、基于支持向量机方法的纳西东巴象形文字识别准确率比已有的采用粗网格特征和基于距离法的识别准确率更高,稳定性更好,可以在此基础上继续进行东巴象形文字的识别研究.

     

    Abstract: Character recognition is an important application area of pattern recognition.Naxi Dongba pictographs are the world's only living pictographs and there is little study on their recognition.Two kinds of characteristics are firstly extracted from the characters with one by the coarse grid method and the other by the topological method.Then the support vector machine (SVM) is used to identify those characters with the two kinds of characteristics compared with the nearest neighbor method.Experiments are designed and the results show that SVM method using the combination of two kinds of characteristics performs better than other methods.

     

/

返回文章
返回