节点文献

一种身份证扫描图像的二值化方法

A Binarization Method to Process Scanned Images of ID Card

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 姚玉花成瑜

【Author】 YAO Yu-hua, CHENG Yu (College of Information Science and Technology, Nanjing University of Aeronautics and Astronautics, JS Nanjing 210016,China)

【机构】 南京航空航天大学信息科学与技术学院南京航空航天大学信息科学与技术学院 江苏南京210016江苏南京210016

【摘要】 本文提出了一种兼顾整体灰度水平和像素邻域特征的身份证扫描图像的二值化方法——嵌入式多阈值动态自适应的二值化方法。该算法模型简单 ,自适应强 ,处理速度快 ;从实验结果上看 ,能滤除噪声 ,而且不出现字符笔划粘连、断裂现象。

【Abstract】 This paper has proposed a new image binarization method to process scanned images of ID card, which is named as Dynamic Self-adaptive Binarization with Embedded Multi Thresholds. This approach can take into account the gray levels of the whole frame and the neighborhood feature of the pixels simultaneously. With this method, image binarization is easy, fast and self-adaptive. Experiment results show that this approach has the advantage of getting rid of adhesion and rupture of characters in scanned ID cards, as well as good cancellation of the noise.

【关键词】 身份证二值化邻域特征
【Key words】 ID cardbinarizationneighborhood features
  • 【文献出处】 机械制造与自动化 ,Jiangsu Machine Building & Automation , 编辑部邮箱 ,2002年05期
  • 【分类号】TP391.41
  • 【被引频次】6
  • 【下载频次】140
节点文献中: 

本文链接的文献网络图示:

本文的引文网络