摘要
介绍了一种利用离散小波变换提取特征的二级签名认证方法。第一级认证基于签名的坐标曲线,第二级认证基于签名灰度图像的直方图,分别获得有关签名曲线和签名图像的小波特征。严格限制第一级认证的错误率,使得第一级认证成功后不再使用第二级认证;否则,将使用第二级认证。实验结果表明,与每次认证都使用两级的认证模式相比,该模式可以有效地缩短签名认证的平均时间。
A kind of method of two - level signature verification based on discrete wavelet transform (DWT) was introduced. The first level verification is based on the signature drawing curves, and the second one is based on the histogram, of signature image. Using the DWT, the wavelet features of the signature drawing curves and signature image are obtained respectively. The strict limitation to the error rate of the first level verification makes the second level verification unnecessary after the success of the verification of the first level. The results of the experiment indicate that compared with the two level verification patterns which were used each time, this pattern can efficiently decrease the computational time needed in signature verification.
出处
《武汉理工大学学报(信息与管理工程版)》
CAS
2008年第1期31-34,共4页
Journal of Wuhan University of Technology:Information & Management Engineering
基金
教育部"十一五"科研规划资助项目(2006JKS2038)
关键词
签名认证
小波变换
动态时间弯曲
signature verification
wavelet transform
dynamic time wrapping