一种脱机手写签名认证方法  

An off-line handwritten signature authentication method

作　　者：朱浩悦[1] 

机构地区：[1]西安文理学院计算机科学与技术系,陕西西安710065

出　　处：《电子设计工程》2009年第9期39-41,共3页Electronic Design Engineering

基　　金：西安文理学院专项科研基金资助项目(KYC200738)

摘　　要：研究脱机签名认证的特征抽取和比较决策。针对脱机手写签名验证特点,提出一种主成分特征提取和径向基神经网络相结合的脱机手写签名认证方法。首先使用主成分特征提取算法对签名图像进行特征提取和降维处理,然后基于主成分特征设计径向基神经网络识别签名的真伪。实验结果表明该方法可取得较好的识别效果。This paper makes further research on feature extraction and decision-making of off-line handwritten signature verification.Aiming at the characteristic of off-line handwritten signature recognition,an off-line handwritten signature au- thentication method using principal character extraction and radial basis neural network is presented.Firstly,principal char- acter extraction algorithm is applied to extract feature and reduce dimension.Then,based on principal character extraction algorithm,an radial basis neural network is designed to identify the authenticity of signatures.The experiment alresuhs show that this method can obtain better identification effect.

关 键 词：脱机手写签名认证 生物测定 主成分特征提取 径向基函数神经网络 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]