期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2006
卷号：6
期号：2B
页码：135-135~138
出版社：International Journal of Computer Science and Network Security
摘要：Multimodal biometric fusion and identity authentication technique help achieve an increase in performance of identity authentication system. In this paper, combined with the characteristics of 2-Dimensional Fisher Linear Discriminant Analysis (2DFLD), a method for face and iris feature fusion is presented on the basis of feature extraction and fusion. Experimental results on ORL (Olivetti Research Laboratory) face database and CASIA (Chinese Academy of Sciences, Institute of Automation) iris database show that, the problem that within-class scatter matrix maybe singular can be solved by the method, and a higher correct recognition rate can be gained. Results also demonstrate the validity of the method, and a new approach is supplied for multimodal biometric feature fusion and recognition.
