一个基于神经网络的动态手写签名验证模型

手写签名验证作为一种有效的生物身份认证技术,具有广泛的应用前景,但由于签名易变化的特点,其性能还不够理想。最近神经网络越来越多地在模式识别领域使用并取得了很好的效果,但在签名验证中尚不多见。本文提出并实现了一个基于BP神经网络的动态手写签名验证原型,并通过实验对网络结构进行了分析。通过对从10人收集的190个本人签名和371个伪造签名评价EER=2.16％,还是比较满意的。

动态手写签名验证技术概述

With the development of e-commerce, people are attaching more importance to biometrics authenticationtechniques. Dynamic handwritten signature verification is an effective one. This paper introduces the basic process ofdynamic handwritten signature verification, some algorithms in use and performance evaluation.

一个在线手写签名身份认证系统的设计与实现

With the development of computer network, Biometrics, as a kind of information security techniques, is thought more and more of. Handwritten signature verification is an effective one. This paper introduces the design and implementation of an on-line handwritten signature verification system which is based on dynamic programming matching algorithm and give a brief evaluation of its performance. The performance of the system is satisfying according to the experiment.

一个基于HMM的动态手写签名认证系统的设计与实现

动态手写签名认证技术作为一种有效的生物认证技术,具有不易遗忘、模仿困难和数据获取简单自然等优点,本文介绍了一个基于隐马尔可夫模型(HMM)的动态手写签名认证系统的设计与实现,该系统将签名认证和识别相结合,简化了认证过程.同时能自动学习,适应用户书写习惯的变化,从实验数据来看,其性能是较满意的.