| 摘要: |
| 提出了曲线化汉字特征的概念,讨论了3类可曲线化的汉字统计特征.利用P形傅里叶算子,对这些曲线化特征作傅里叶变换,可提取到最终的识别特征.实验结果表明,该方法在手写体汉字识别领域具有良好的应用前景,尤其适用于细分类. |
| 关键词: 手写体,汉字识别,统计特征,细分类,傅里叶变换. |
| DOI： |
| 分类号: |
| 基金项目: |
|
| Application of P-Fourier Transform to Handwritten Chinese Character Recognition |
|
YAO Dan-lin,CHEN Huo-wang,YIN Jian-ping
|
| Abstract: |
| In this paper, the conception of Chinese character's corresponding curve feature is proposed, and three kinds of statistical features of Chinese characters, which can be represented by curve, are discussed. Applying P-Fourier operator to them, the final features for character recognition are extracted. Experimental results indicate that the proposed method is very promising for the recognition of handwritten Chinese characters, and is especially suitable for the fine classification. |
| Key words: Handwritten character, Chinese character recognition, statistical feature, fine classification, Fourier transform. |
