摘要
为降低阅卷的人力成本,提高考试情况分析的准确率和效率,对答题卡模板生成以及答题卡识别的算法进行研究和设计。实现动态设置答题卡信息,用word文档形式生成答题卡模板,以图片形式读入实物答题卡,通过图像识别获取答题卡填涂内容,将识别结果与设定的答案信息进行比对,得出考试结果。系统运用OpenCV计算机视觉库对答题卡图片进行灰度化、降噪、模板匹配、透视变换、二值化等一系列图像处理,最终实现答题卡模板生成以及答题卡识别。
In order to reduce the labor costs of marking,improve the accuracy and efficiency of the analysis of the examination,the algorithm of answer sheet template generation and answer sheet recognition is researched and designed.First,the parameters of an⁃swer sheet are set dynamically,the answer sheet template is generated in word document form,the physical answer sheet is read in picture,then,the contents of the answer sheet will be obtained by image recognition,the recognition result and the right answer will be compared,at last the test result will be obtained.The system uses the OpenCV computer vision library to perform a series of image processing such as graying,noise reduction,tem⁃plate matching,perspective transformation,binarization,etc.on the answer card pictures,finally,the answer sheet template is gen⁃erated and answer sheet recognition are realized.
作者
韦溢辉
刘汉英
WEI Yi-hui;LIU Han-ying(Guilin University of Technology,Guilin 541004,China)
出处
《电脑知识与技术》
2020年第29期23-27,共5页
Computer Knowledge and Technology
基金
2016年广西高校“嵌入式技术与智能信息处理”重点实验室基金项目(项目编号:2016-02-06)
2020年度广西高等教育本科教学改革项目(项目编号:2020JGB211)。
关键词
答题卡生成
答题卡识别
OpenCV图像处理
模板匹配
answer sheet generation
answer sheet recognition
OpenCV image processing
template matching
