基于自定义模板的OCR技术及应用  

OCR Technology and Application Based on Custom Template

作　　者：章倩 王梓祺 ZHANG Qian;WANG Ziqi(The 28th Research Institute of China Electronics Technology Group Corporation,Nanjing 210023,China;Command and Control Engineering College,Army Engineering University of PLA,Nanjing 210007,China)

机构地区：[1]中国电子科技集团公司第二十八研究所,南京210023 [2]陆军工程大学指挥控制工程学院,南京210007

出　　处：《指挥信息系统与技术》2023年第5期94-98,共5页Command Information System and Technology

摘　　要：信息资料电子化工作中,基于深度学习的光学字符识别(OCR)技术虽然是一种有效的文字识别方法,但也存在人工干预多以及结构化数据识别困难等问题。为了提高应用系统自动化程度以及满足用户定制化需求,提出了一种基于自定义模板的OCR技术。实际应用表明,基于自定义模板的OCR技术既可减少人工干预又可解决结构化数据识别问题,有助于提高信息加工与处理效率。In the job of electronic information,although the optical character recognition(OCR)tech-nology based on deep learning is an effective method for text recognition,it also has many problems such as excessive manual intervention and difficult recognition in recognizing structural dada.In order to improve the automation degree of application system and meet the needs of user customization,an OCR technology based on custom template is proposed.Practical application show that the OCR tech-nology based on custom template can reduce the manual intervention,and also can solve the problem of structural data recognition.Thus,it is helpful to improve the efficiency of information processing and processing.

关 键 词：光学字符识别技术 深度学习 自定义模板 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]