计算机工程与应用 ›› 2022, Vol. 58 ›› Issue (2): 27-42.DOI: 10.3778/j.issn.1002-8331.2107-0207

• 热点与综述 • 上一篇    下一篇

手写汉字评价方法研究进展

肖雪,李成城   

  1. 内蒙古师范大学 计算机科学技术学院,呼和浩特 010022
  • 出版日期:2022-01-15 发布日期:2022-01-18

Research Progress on Evaluation Methods of Handwritten Chinese Characters

XIAO Xue, LI Chengcheng   

  1. College of Computer Science and Technology, Inner Mongolia Normal University, Hohhot 010022, China
  • Online:2022-01-15 Published:2022-01-18

摘要: 手写汉字评价推动了计算机辅助教学的快速发展,如何通过手写汉字评价技术,在无教师帮助的情况下实现手写汉字的等级/规范性评价是当前研究的重点。对手写汉字评价相关概念以及发展趋势进行阐述;对手写汉字评价的不同研究方法进行详细介绍,包括基于规则、特征相似度计算、模糊矩阵以及机器学习等方面,并对各种方法的优缺点进行总结归纳;对手写汉字评价的反馈形式进行介绍,包括数据到文本生成、字形匹配与图形辅助等方面;分析手写汉字评价面临的多个问题,进一步思考其未来的发展。

关键词: 书写质量评价, 特征提取, 特征匹配, 深度学习

Abstract: Handwritten Chinese character evaluation promotes the rapid development of computer-aided teaching. How to realize the grade/normative evaluation of handwritten Chinese characters without teachers’ help through handwritten Chinese character evaluation technology is the focus of current research. Firstly, this paper expounds the related concepts and development trend of handwritten Chinese character evaluation. Secondly, different research methods of handwritten Chinese character evaluation are introduced in detail, including rule-based, feature similarity calculation, fuzzy matrix and machine learning, and the advantages and disadvantages of various methods are summarized. Then it introduces the feedback form of handwritten Chinese character evaluation, including data to text generation, font matching and graphic assistance. Finally, it analyzes many problems faced by the evaluation of handwritten Chinese characters, and further considers its future development.

Key words: writing quality evaluation, feature extraction, feature matching, deep learning