Research on quantitative evaluation of standard Chinese characters written by pen and paper based on neural network
WANG Min1， MA Wan2， ZHU Chuang2， SHI Shanfei2， SHU Jiangbo2， LU Shuaicheng2
(1.Jiangsu Province Support Software Engineering R&D Center for Modern Information Technology Application in Enterprise， Suzhou 215104，Jiangsu， China；2.National Engineering Research Center of Educational Big Data， Central China Normal University， Wuhan 430079， China)
Abstract：Chinese character writing education is an important part of our country's quality education. Currently， there are two major problems in the education of standard Chinese character writing for primary and secondary school students in China. Firstly， there is a lack of effective guidance， with the emphasis being mainly on qualitative descriptions of standard Chinese character writing essentials， and the quantitative characteristics of details are not reflected， making it difficult for students to accurately grasp the writing details； secondly， there is a lack of effective evaluation， with most teachers or parents only able to give an overall qualitative evaluation of students' writing， and unable to provide precise detailed evaluation. To address these issues， this study focuses on the quantitative evaluation of standard Chinese character handwriting using pen and paper in natural writing scenarios， including three main areas of work： 1) using the skeleton image of handwritten Chinese characters and pre-evaluation results to construct a classification evaluation model based on convolutional neural networks， achieving automatic classification evaluation of the standardization of handwritten Chinese characters； 2) using the stroke attribute data of handwritten Chinese characters to achieve quantitative evaluation of writing details and providing suggestions for writing adjustments； 3) using the standardization score of various writing rules of handwritten Chinese characters to construct a writing scoring model based on multiple linear regression， achieving an evaluation of the overall quality of handwritten Chinese characters.
王 敏,马 万,祝 闯,史善飞,舒江波,卢帅成. 基于神经网络的纸笔手写规范汉字量化评价研究[J]. 华中师范大学学报(自然科学版), 2023, 57(6): 813-820.
WANG Min,MA Wan,ZHU Chuang,SHI Shanfei,SHU Jiangbo,LU Shuaicheng. Research on quantitative evaluation of standard Chinese characters written by pen and paper based on neural network. journal1, 2023, 57(6): 813-820.
2023, Vol. 57 
