结合YOL0与Android平台的答题卡自动评分系统的应用研究

打开文本图片集

中图分类号：TP391.4 文献标识码：A 文章编号：2096-4706（2025）15-0144-06

（南宁理工学院，广西桂林541006)

Research on the Application of the Answer Sheet Automatic Scoring System Based on YOLO and Android Platform

MO Nianliang (Nanning College of Technology,Guilin 541006,China)

Abstract:The traditional manual marking method is ineffcient and error-prone.Therefore,anautomatic scoringsystem for answersheets is designedand implemented based onYOLOobject detectionalgorithmand Android platform.Firstly,bybuilding a dep learning environment andpreparingadataset,LabelImg is usedtolabel images,andtheYOLOv5 model is tainedusing thePyTrch frameworkandconverted intoaformatsuitableformobiledevices.Then,anAndroidaplicationbasedonKotlin languageandJetackComposeiseveloed,hichimplementstefutionsofiageacqusitionjectetectiondoing Theexperimentalresultsshowthatthe systemhashighaccuracyandreal-time performance,whichcaneffectively improve the efficiency of marking and provide technical support for educational informatization.

KeyWords:YOLO; Android platform; automatic answer sheet scoring; Object Detection; Deep Learning; PyTorch

0 引言

随着教育信息化的快速发展，考试形式日益多样化，自动化阅卷的需求日益迫切。（剩余12024字）