Computer Engineering and Applications ›› 2017, Vol. 53 ›› Issue (10): 246-250.DOI: 10.3778/j.issn.1002-8331.1512-0030

Previous Articles     Next Articles

Handwritten Chinese character recognition system based on neural network convolution depth

YAN Xiliang1, WANG Liming2   

  1. 1.Zhengzhou University of Industrial Technology, Xinzheng, Henan 451150, China
    2.School of Information Engineering, Zhengzhou University, Zhengzhou 450001, China
  • Online:2017-05-15 Published:2017-05-31

卷积深度神经网络的手写汉字识别系统

闫喜亮1,王黎明2   

  1. 1.郑州工业应用技术学院,河南 新郑 451150
    2.郑州大学 信息工程学院,郑州 450001

Abstract: For the feature extraction methods’restrictions problem of handwritten Chinese character recognition in traditional two hand written Chinese character recognition system, the paper proposes an identification system method which uses convolution neural network to automatically learn Chinese characters similar characteristics. The method uses data from large handwritten cloud platform to train the model, generating similar frequency statistics based on a subset of, and further improve the recognition rate. Experimental results show that the recognition rate with respect to the traditional gradient-based feature support vector machine and nearest neighbor classifier method, this method has improved to some extent.

Key words: handwritten Chinese character, automatic learning, convolution neural network, cloud platform, recognition rate

摘要:

针对传统两级手写汉字识别系统中手写汉字识别的特征提取方法的限制问题,提出了一种采用卷积神经网对相似汉字自动学习有效特征进行识别的系统方法。该方法采用来自手写云平台上的大数据来训练模型,基于频度统计生成相似子集,进一步提高识别率。实验表明,相对于传统的基于梯度特征的支持向量机和最近邻分类器方法,该方法的识别率有一定的提高。

关键词: 手写汉字, 自动学习, 卷积神经网, 云平台, 识别率