• Acta Photonica Sinica
  • Vol. 51, Issue 11, 1111001 (2022)
Ruyu YAN1, Xiaoxia WANG1、*, Jiangtao XI1、2、**, Fengbao YANG1, and Daerhan BAO3
Author Affiliations
  • 1School of Information and Communication Engineering,North University of China,Taiyuan 030051,China
  • 2School of Electrical Computer and Telecommunications Engineering,University of Wollongong,Wollongong NSW 2522,Australia
  • 3Xi′an Microelectronice Technology Institute,Xi′an710054,China
  • show less
    DOI: 10.3788/gzxb20225111.1111001 Cite this Article
    Ruyu YAN, Xiaoxia WANG, Jiangtao XI, Fengbao YANG, Daerhan BAO. Handwritten Font Classification Method Based on Ghost Imaging[J]. Acta Photonica Sinica, 2022, 51(11): 1111001 Copy Citation Text show less
    The realization process of automatic recognition method of handwritten digit
    Fig. 1. The realization process of automatic recognition method of handwritten digit
    CNN neural network architecture model
    Fig. 2. CNN neural network architecture model
    The accuracy of digit from 0 to 9 under different sampling rates of CNN and DNN models
    Fig. 3. The accuracy of digit from 0 to 9 under different sampling rates of CNN and DNN models
    Accuracy rate of 10 letters under different sampling rates and CNN model precision,recall rate,F1 score
    Fig. 4. Accuracy rate of 10 letters under different sampling rates and CNN model precision,recall rate,F1 score
    The loss error curve of different models using the cross-entropy loss function
    Fig. 5. The loss error curve of different models using the cross-entropy loss function
    ModelLowest accuracyHighest accuracyAverage accuracy
    CNN95.65%98.47%97.65%
    DNN87.56%92.99%92.34%
    Table 1. Comparison of accuracy based on CNN and DNN models
    ParameterDNNCNNLift value
    Precision86.50%97.25%10.75%
    Recall rate86.40%98.03%11.63%
    F1 score86.31%97.60%11.09%
    Table 2. Comparative analysis of precision,recall rate and F1 score under DNN and CNN models
    Ruyu YAN, Xiaoxia WANG, Jiangtao XI, Fengbao YANG, Daerhan BAO. Handwritten Font Classification Method Based on Ghost Imaging[J]. Acta Photonica Sinica, 2022, 51(11): 1111001
    Download Citation