基于欧氏权距离法的多生物特征身份鉴别

通过融合不同的生物特征可以改进身份鉴别系统的验证性能。本文借鉴多传感器信息融合技术中聚类分析的思想,提出了基于欧氏权距离法的多生物特征融合的身份鉴别方法。该方法并不局限于特定的两类或多类生物特征,同时该方法克服了传统的估计方法和统计方法需要先验概率和概率密度函数的缺点,更加符合实际应用的需要。

Multi-Modal Data Fusion for Person Authentication based on Weighted Euclidean Distance Method

The use of multiple modalities can improve the accuracy and robustness of person authentication. The multi- modal data fu-sion for person authentication based on Euclidean distance- weighted method is proposed in this paper according to cluster analysis method used in multi- sensor data fusion system. The method is not limited in two or more special bio- features. Compared with the traditional estimate method and Statistic method, the proposed method overcomes the limitations which need prior probabilistic model and probability density function. The method is more effective in practice.