答题卡识别matlab源码,基于图像的答题卡识别与成绩统计（matlab版）.docx

基于图像的答题卡识别与成绩统计(matlab版)

中文题目：基于图像的答题卡的判定与成绩统计外文题目：The Judgment and Performance Statistics based on Image for Answer Card毕业设计(论文)共 62 页(其中：外文文献及译文15页) 图纸共0张完成日期 2017年6月 答辩日期 2017年6月摘要现代教育随着科技和计算机技术的发展而发生着巨大的变化，考试的公平公正公开性也越来越受到重视，而在竞争如此巨大的今天，正确的考试成绩对于学生、家长、老师以及学校都非常的重要，学生和老师可以根据考试成绩查漏补缺及时调整学习和教学方案，学校通过成绩可以加强教学监督和调整教学计划。在其他方面，正确的考试成绩也是选拔人才及其重要的方法之一。本文使用Matlab对国内某通用型号的答题卡进行了识别和成绩统计，处理的步骤可分为预处理、平滑处理、灰度化、二值化、hough变换、图像旋转、形态学滤波、区域划分、坐标区域分割、成绩统计和GUI功能封装；其中，用到的主要技术是图像放缩、二维高斯卷积滤波器使用、基于Otsu算法的全局阈值计算、基于hough变换的直线检测、有效元素标记与提取、利用坐标对图像的区域划分。最终使用GUI界面将功能进行整合，实现单张答题卡的手动识别与多张答题卡的自动识别，并能完成成绩统计。关键词：图像处理；答题卡；hough变换；成绩统计AbstractModern education with the development of science and technology and computer technology has undergone tremendous changes,fair and open examination of the examination more and more attention, and in the competition so great today, the correct test scores for students, parents, teachers and schools are very important,students and teachers can adjust the learning and teaching programs according to the examination results. The school can strengthen the teaching supervision and adjust the teaching plan through the achievements. In other respects, the correct test scores are also one of the important ways to select talent and its important.In this paper, we use Matlab to identify and score statistics on a common model of domestic models. The processing steps can be divided into preprocessing, smoothing, grayscale, binarization, hough transform, image rotation, morphological filtering, , The coordinates of the regional segmentation, performance statistics and GUI function package; which, the main technology used is image scaling, two-dimensional Gaussian convolution filter, based on Otsu algorithm global threshold calculation, based on hough transform linear detection, effective elements Marking and extracting, using coordinates to divide the area of the image. The final use of the GUI interface will be functional integration, to achieve a single answer card manual identification and multiple