J4

• • 上一篇    下一篇

在线手写签名组合最相似笔段的识别算法

张达治,马驷良,王彦   

  1. (吉林大学数学学院,长春130012)
  • 收稿日期:2004-12-15 修回日期:1900-01-01 出版日期:2005-09-26 发布日期:2005-09-26
  • 通讯作者: 马驷良

Recognition Algorithm of the most Similar Strokes of Online Handwritten Signature

ZHANG Da-zhi,,MA Si-liang,WANG Yan-gang   

  1. (College of Mathematics,Jilin University,Changchun130012,China)
  • Received:2004-12-15 Revised:1900-01-01 Online:2005-09-26 Published:2005-09-26
  • Contact: MA Si-liang

摘要: 在线手写签名识别是通过人的生物特征进行身份验证的技术.提出一种基于笔画样本库的识别方法,避免了动态特征点提取、时序段匹配等方法由于书写环境和人物即时的生理特性所带来的问题,同时针对每一笔画的稳定性提出了加权距离计算公式.实验证明该方法行之有效,在保证高识别率的同时又降低了误拒率.

关键词: 在线手写签名识别, 笔段, 误拒率

Abstract: Online handwritten signature recognition is the technology by which the identity is verified via person’s biological characteristic.This paper puts forward a recognition algorithm based on stroke sample database,which prevents the questions of picking-up dynamic characteristic method and matching time sequence section method etc due to writing environment and person’s instant physiological characteristic. At the same time this paper puts forward a weighted distance calculation formula based on weighting to the stability of every stroke.The experiment has proved that this method is effectual.This method reduces the false reject rate (FRR)while guaranteeing high recognition rate.

Key words: online handwritten signature recognition, stroke, false reject rate

中图分类号: 

  • TP391