基于CNN的文字识别系统设计  

CNN-based Text Recognition System Design

作　　者：田芙荣 刘静静(指导)[1] 陈方苇 TIAN Furong;LIU Jingjing;CHEN Fangwei(Information Science and Technology School of Zhejiang Shuren University,Hangzhou,Zhejiang,310015,China)

机构地区：[1]浙江树人大学信息科技学院,浙江杭州310015

出　　处：《浙江树人大学学报（自然科学版）》2019年第3期12-17,共6页Journal of Zhejiang Shuren University(Acta Scientiarum Naturalium)

基　　金：2018年度浙江省大学生科技创新项目(2018R421023).

摘　　要：文字识别在人们日常生活中的应用非常广泛,但因纸质文件转化为文档图片过程中经常不可避免地产生一些问题,一定程度上给文字识别造成了困难.文章围绕文字识别问题,针对提高文字识别率及识别速率,利用OpenCV技术实现图像的校正,并通过深度学习技术和OCR技术,研发文字识别系统.该系统研究成果应用广泛,尤其在数字化和自动化领域,可促进办公自动化的进一步发展.Text recognition is widely used in people’s daily life.However,in the process of transforming paper documents into document pictures,there often exist some inevitable problems,which to some extent cause difficulties in text recognition.This paper focuses on the problem of text recognition,and aims at improving the rate and speed of text recognition.OpenCV technology is used to realize image correction,and text recognition system is developed through deep learning technology and OCR technology.The research results from the system can be widely used,especially in the field of digitization and automation,and can promote further development of office automation.

关 键 词：文字识别 图片校正 卷积层 深度学习 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论] TP183[自动化与计算机技术—计算机科学与技术]