图像处理耦合模板定位的答题卡识别研究与应用.pdfVIP

计算技术与自动化 图像处理耦合模板定位的答题卡识别研究与应用 郝平 （陕西工业职业技术学院，信息工程学院，陕西，咸阳，712000 ） 摘要：现代大型考试中已广泛使用答题卡填涂方式，要求机器阅卷，从而提高阅卷速度和保证客观性。而当前大多数机器阅 卷中采用的识别算法基于模糊识别，即针对某类型的试卷，更换多种试卷或者同种试卷不同采集方式下很难准确对应，具有一 定的局限性。对此，本文提出了一个基于OpenCV 耦合模板定位的答题卡识别机制。首先基于人机交互划定学号区与客观题区； 然后基于图像处理算法定位得到填涂位置，评价填涂结果，完成答题卡识别。本系统模板制作模块由 C#编程实现，答题卡识 别由C++和OpenCV 实现。最后测试了本文机制性能，结果表明：与基于模糊识别的普通方法相比，本文机制具有更好的定位 效果和识别准确度。 关键词：模板定位；OpenCV ；人机交互；模糊识别；图像处理 中图分类号：TP391 文献标识码：A The Research and Application on The Exam Card Reading Based on Image Processing and Template Positioning Hao Ping College of Information Engineering, Shaanxi Polytechnic Institute, Xianyang, Shanxi, 712000, China （ ） Abstract: Filling of answer card has been widely used in modern large-scale tests, requirements of machine scoring, so as to improve the speed and ensure the objectivity of scoring. And the recognition algorithm used in most current machine scoring based on fuzzy identification, in a variety of papers or the same test in different collection methods,it is difficult to accurately correspond, with certain limitations. Therefore, this paper proposes a card recognition mechanism of OpenCV coupling template location based on the answer. Firstly,the delimited student ID and the objective questions district is obtained based on the image processing algorithm; then evaluating fill position results, the filling, completing answer card identification. This system templates creating module is programmed by C#, the answer card recognition is achieved by C++ and OpenCV. The final test the performance of the mechanism, the results show that: compared with general method based on fuzzy recognit