基于小波变换提取特征点的在线手写签名认证算法

提出了一种基于小波变换的认证算法，首先对F-Tablet手写力平台获取的签名书写过程中的三维力信息和二维坐标信息，使用小波变换的方法来提取这些信息中的关键点，得到五维特征序列，使用聚类方法动态生成多个模板，然后通过多数投票方案融合各个特征序列的决策值得到认证结果。在基于F-Tablet平台构建的签名样本库上对此算法进行了实验，平均等错误率为2.83%。实验结果表明，该算法不仅可以减少签名数据存储量，而且能够缩短认证过程的持续时间，增加认证过程的有效性。

On-line handwritten signature verification algorithm based on wavelet transform to extract characteristic points

A verification algorithm based on Wavelet Transform (WT) was proposed in this paper. Firstly, the algorithm used WT to exact characteristic points of the handwritten information including three-dimension force and two-dimension coordinate information that were obtained by the F-Tablet. And then it built five-dimension feature sequences and dynamically created multi-templates by using clustering. Finally, after the fusion of the above-mentioned five-dimension feature sequences, the algorithm computed the verification results by majority voting scheme. This algorithm was applied to a signature database acquired by F-Tablet, and the performance evaluation in even EER (Equal Error Rate) was improved to 2.83%. The experimental results show that the algorithm not only reduces the amount of stored data, but also minimizes the duration of the whole processing during the authentication phase, and increases the efficiency of signature verification.