| 基于多变量多项式模型的多模态生物特征分数层融合算法 |
| |
| 引用本文： | 张志坚,张培仁.基于多变量多项式模型的多模态生物特征分数层融合算法[J].中国科学技术大学学报,2009,39(11). |
| |
| 作者姓名： | 张志坚 张培仁 |
| |
| 作者单位： | 中国科学技术大学自动化系,安徽,合肥,230027 |
| |
| 摘 要： | 提出了一种基于多变量多项式模型的多模态生物特征分数层融合算法(MPMSF),用多变量多项式模型获取融合分数,用最小二乘法确定模型中的参数.并对多变量多项式模型进行简化,以减小计算量.MPMSF的特点在于不需要分数归一化.采用交叉验证的方法,评估了各种分数层融合算法的正确接受率和分数归一化、单模态性能对分数层融合算法的影响.开集测试表明,MPMSF算法优于传统的分数层融合算法.同时,MPMSF满足多模态生物特征识别系统实时性的需求.
|
| 关 键 词： | 多模态生物特征识别 分数层融合 多变量多项式模型 交叉验证开集测试 |
A score fusion algorithm based on the multivariate polynomial model |
| |
| ZHANG Zhi-jian,ZHANG Pei-ren.A score fusion algorithm based on the multivariate polynomial model[J].Journal of University of Science and Technology of China,2009,39(11). |
| |
| Authors: | ZHANG Zhi-jian ZHANG Pei-ren |
| |
| Abstract: | A novel multivariate polynomial model based score fusion algorithm(MPMSF)was proposed.A fused score was obtained from the matching scores of single biometrics by using the multivariate polynomial model.The parameters of the function were estimated using the least squares method(LSM).Also,cross validation was employed for score fusion algorithm evaluation.The open set test shows that MPMSF outperforms other score fusion methods and that its CPU time can meet the real time requirement of multimodal biometrics. |
| |
| Keywords: | multimodal biometric score fusion multivariate polynomial model cross validation open test |
| 本文献已被 万方数据 等数据库收录！ |
