基于边框线的版面分析方法应用于选票处理

民主选举过程中选票的统计是一项严肃而繁琐的工作,传统的手工操作方法不仅效率低,而且在统计过程中很容易出错,影响了选举工作的进度及质量。选票信息的自动处理能提高录入速度、准确率和公正性。选票信息自动处理的核心是选票版面结构获取和填写信息识别。该文提出了一种基于边框线的选票版面分析方法,并采用OCR(OpticalCharacterRecognition)识别填写信息,该方法具有对选票的设计更方便灵活、纸张和填写的要求低、精度高、通用性好的特点。

Method of Vote Information Processing Based on OCR

Vote statistics is a solemn and cockamamie work in democratic e lection,the method of manual vote is not only inefficiency but also prone to m ake a mistake.Vote automatic processing system can assure process speed,accura cy and justness.The vote processing approach consists of vote layout analysis and filling character recognition.A vote layout analysis approach based on bou nding rectangle is been put forward.Optical Character Recognition is introduce d to recognize the filling information in the vote.It is provided with convenie nce in vote design,high accuracy and universal property.