摘要
为全面深入地掌握汉字识别技术的研究现状与应用进展,基于WOS数据库中汉字识别研究领域的文献数据,利用CiteSpace可视化软件进行文献计量与知识图谱分析。结果表明:中国的高校和研究所是汉字识别研究的核心力量;汉字识别技术的发展历程由早期的传统方法OCR技术演化为中期的联机手写汉字识别,再到近期基于深度学习的汉字识别研究;研究前沿为手写汉字识别、混合语言模型、生成对抗网络和场景文本识别;汉字识别技术将朝着高效率及智能化方向发展。
In order to keep a comprehensive and systematic understanding of the progress of Chinese character recognition,the literature data of Chinese character recognition in Web of Science database are selected and knowledge mapping analysis is made by CiteSpace visualization software.The research results show that Chinese universities and research institutes are the core force of Chinese character recognition.The development of Chinese character recognition technology has evolved from the early traditional OCR technology to the online handwritten Chinese character recognition in the middle period,and then to the deep learning in recent research of Chinese character recognition.The research frontiers topics are handwritten Chinese character recognition,hybrid language model,generative adversarial networks and scene text recognition.Chinese character recognition technology will develop towards high efficiency and intelligence.
作者
李阳娟
LI Yangjuan(Engineering Training Center,Putian University,Putian Fujian 351100,China)
出处
《科技和产业》
2022年第4期180-187,共8页
Science Technology and Industry
基金
福建省教育厅中青年科技项目(JAT200515)
莆田学院校内科研项目(2019030)。
关键词
汉字识别
光学字符识别
深度学习
手写体汉字
Chinese character recognition
OCR
deep learning
handwritten Chinese character