选票选举系统中选票图像的预处理方法研究  

Research of Image Pre-Processing in Vote Processing System

作　　者：张婷[1] 吴元君[2] 黄俊[1] 吴建国[1] 

机构地区：[1]安徽大学计算机科学与技术学院,安徽合肥230039 [2]安徽财贸职业学院,安徽合肥230601

出　　处：《计算机技术与发展》2007年第4期225-228,共4页Computer Technology and Development

摘　　要：基于图像识别技术的选举计票系统是由图像获取系统、机械传动系统、计算机、网络等构成的光、机、电、信息一体化装置,采用与普通光学标记阅读机完全不同的技术,使用光学字符识别的方法、算法完成选票上标记的识别。此系统中图像识别是识别出与会代表填写的表示同意、反对或弃权的“√、×、○、●、\、/”等各种手写符号,它属于脱机手写体汉字识别。重点讨论了图像灰度化、二值化、倾斜校正、平滑以及细化等选票信息识别中非常重要的图像处理技术。针对选票图像的特性,以提高识别率为目的,提出了恰当的预处理方法。Vote processing system based on image recognition is an integrative device composed of imaging collection, transmission mechanism, computer and networking subsystems. Differs from other vote processing system which is based on the optical mark recognition, the system adopts methods, algorithms of optical character recognition techniques to recognize the handwritten symbols on votes. So this systerm is to recognize all kinds of handwritten symbols written by candidates such as＂√,× ,○,●, ／ ,/＂ ,which belongs to off- line handwritten Chinese character recognition. In this paper, several important image processing techniques of vote information recognition are discussed, such as transforming of the color plate to gray plate, image binarization, skew correction, image smoothing, thinning, etc. In order to improve the rate of correct identification, put forward proper methods of image pre- processing.

关 键 词：选举系统 图像识别 预处理 脱机手写体汉字 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]