文献类型：期刊文献

中文题名：基于手写数字识别的试卷分析系统

英文题名：System of Test Paper Analysis Based on Handwritten Digit Recognition

作者：闫超超[1];耿维忠[1];李华志[1];康家豪[1]

第一作者：闫超超

机构：[1]新乡学院计算机与信息工程学院,河南新乡453003

第一机构：新乡学院计算机与信息工程学院

年份：2023

期号：3

起止页码：50-52

中文期刊名：机械工程与自动化

外文期刊名：Mechanical Engineering & Automation

基金：河南省科技攻关计划项目(182102311123);新乡学院校级课程思政示范课程项目(4101821205)。

语种：中文

中文关键词：手写数字识别;试卷分析系统;卷积神经网络;OpenPyXL;Python-Docx

外文关键词：handwritten digit recognition;test paper analysis system;convolutional neural network;OpenPyXL;Python-Docx

摘要：针对现有试卷分析过程中需要手工录入、统计和分析成绩,导致繁琐费时等问题,提出了基于手写数字识别的试卷分析系统。首先利用图像处理技术增强需处理的试卷成绩表区域;然后使用改进的LeNet-5卷积神经网络识别分数;最后,采用OpenPyXL和Python-Docx库函数完成成绩保存、分析和报告自动生成。实验表明,该系统能提高试卷分析效率,减轻教师工作负担。
In view of the tedious and time-consuming problems caused by manual input,statistics and analysis of scores in the process of test paper analysis,a test paper analysis system based on handwritten digit recognition is proposed.First of all,the image processing technology is used to enhance the area where the test paper score sheet needs to be processed.Then the improved LeNet-5 convolution neural network is used to identify the score.Finally,OpenPyXL and Python-Docx library functions are used to save,analyze and report the results automatically.The experimental results show that the system can improve the efficiency of test paper analysis and reduce the burden of teachers.