【作者】 郭震；

【导师】 郑建彬；

【作者基本信息】 武汉理工大学 ， 通信与信息系统， 2006， 硕士

【摘要】 传统的身份鉴别方法(如钥匙、口令等)已不能满足社会的需要，基于生物特征的身份鉴别技术的发展为我们提供了一种更加方便和可靠的解决方案。签名作为一种行为特征，相比其它生物特征有着更易于获取和能够共享的优点，是身份鉴别中应用最为广泛的生物特征之一。 演化计算是目前数学上应用得较为广泛的一种算法，与传统的算法相比，其最主要的特点在于演化计算具有智能性和本质并行性的特征。通过算法的搜索、加速，可以自行找到所求问题的最优解。 本文将演化计算和在线签名验证结合起来，提出了基于演化计算的签名验证方案，方案分为两个部分，即签名验证曲线段的演化计算程序和签名曲线的动态分割。 签名验证的演化计算程序的基本思路是通过演化计算找到波形的平移、伸缩分量的取值，从而得到参考签名段和测试签名段的最小距离，通过比较这个最小距离与预先设定的匹配门限的大小来判断测试签名段是否与参考签名段匹配。根据这个思路，本文给出了签名验证的演化计算程序的数学模型，并在数学模型的基础上，给出了签名验证的演化计算程序的具体算法。 签名曲线段的动态分割是签名验证的演化计算程序的前提，只有对签名曲线段进行了正确的分割，才能得到签名验证的准确结果。本文首先比较了几种传统的签名曲线分割方法，指出了签名曲线段的动态分割的必要性，然后详细的介绍了这种动态分割的算法。 文章的最后给出了基于演化计算的签名验证方案的实验结果，并且对实验结果进行了分析，给出了方案中各个参数的恰当的取值，并计算出了该方案的误纳率(FAR)和误拒率(FRR)。 实验的结果表明，这种验证方法达到了较高的验证性能，具有很好的实用价值及应用前景。更多还原

【Abstract】 Traditional methods for verification and recognition fail to satisfy the society demands. Fortunately, development of the techniques for recognition and verification of identification based on biometric features provides a more convenient and more reliable solution. As a kind of behavioral feature, signature has advantages of easy acquirement and sharing. It turns out one of the most popular features in identity recognition and verification.Evolutionary Computation(EC) is an algorithm which is widely used in Mathematics nowadays. Compared with the conventional algorithms, the main specialty of EC lies in its intelligence and characteristic parallelism. It can find the optimal solutions to the problem by itself through the search and acceleration.This paper combines EC with signature verification and proposes a scheme which is signature verification based on EC. This scheme divided into two parts, namely, the evolutionary computation program of signature curve verification and the dynamic segmentation to signature.The basic idea of the EC programming of signature verification is to find the values of scale factor and shift of signature curve which can calculate the minimum distance between reference signature and test signature. By compared with the minimum distance and the predetermined threshold, we can determine whether the section of test signature matches the section of reference signature. According to this thought, this paper presents a signature verification mathematical model of EC program, and proposes its detail algorithm.The dynamic segmentation is the premise of signature verification based on EC. Only when we make the correct segementation to the signature curve, we can get the accurate result of signature verification. At first, this paper compares several traditional segmentation methods of signature curve and points out the necessarity of the dynamic segmentation. Then, the paper explains the dynamic segmentation algorithm in detail.Finally, this article presents the experimental results of signature verification based on EC, and analyzes these results, points out the proper value of various parameters in program. At the end of this paper, we calculate the FAR and FRR of this scheme.The experimental results show that this way of signature verification achieves a更多还原