无定位点答题卡扫描图像纠偏算法的研究与设计

Research and design of de-skew algorithm for scanned images of answer sheet with no anchor point

答题卡扫描图像纠偏处理是自动化阅卷系统中的重要环节。使用无定位点答题卡可以降低阅卷的成本,有利于自动化阅卷系统在中小规模考试中推行。为了完善答题卡扫描图像纠偏处理的过程,文章提出了一种适用于无定位点答题卡扫描图像纠偏的算法。本算法利用了无定位点答题卡中的设计元素和排版规律进行辅助纠偏。经测试此算法有着较好的纠偏成功率。

Skew correction for scanned image of answer sheet is an important part of automatic paper marking system. The use of no anchor point answer sheet can reduce the cost of automatic paper marking system, which will generalize the use of automatic paper marking system in small and mid-sized exam. In order to improve the de-skew algorithm for the scanned image of answer sheet, this paper presents a de-skew algorithm for scanned images of answer sheet with no anchor point. This algorithm makes good use of the design elements and typesetting rules in the answer sheet with no anchor point. The test results show that this algorithm achieves good performance in skew correction.