【摘要】 为降低成本与实现答题卡的自动化识别，设计了基于机器视觉的答题卡识别系统，对图像提取、区域划分及识别算法分别进行研究。通过形态学操作、轮廓检测、仿射变换等技术，实现对答题卡图像的提取；基于最大矩形框检测算法，实现答题卡的区域划分；利用透明通道(Alpha通道)对标准答案与待识别答案填涂区域进行抠图，将处理后的图像匹配叠加；利用叠加图像的RGBA颜色空间特性、同步头的灰度投影结果与指针算法共同完成答题卡的识别。试验结果表明，系统的使用与维护成本低、操作方便且具有较好识别效果。更多还原

【Abstract】 In order to reduce the cost and realize the automatic identification of answer sheets, an answer sheet identification system based on machine vision is designed, and the image extraction, area division and identification algorithms are respectively studied. Firstly, through morphological operation, contour detection, affine transformation and other technologies, the image of the answer sheet is extracted. Based on the largest rectangular box detection algorithm, the area division of the answer sheet is realized. Secondly, the filling area of the standard answer and the answer to be recognized are cut out by using the transparent channel(Alpha channel),and the processed images are matched and superimposed. Finally, the RGBA color space characteristics of the superimposed image, the grayscale projection result of the sync head and the pointer algorithm are used to complete the identification of the answer sheet. The experimental results show that the system has low cost of use and maintenance, convenient operation and good recognition effect.更多还原