期刊名称：BVICAM's International Journal of Information Technology
印刷版ISSN：0973-5658
出版年度：2009
卷号：1
期号：1
出版社：Bharati Vidyapeeth's Institute of Computer Applications and Management
摘要：Linear Discriminant analysis is a commonly used and valuable approach for feature extraction in face recognition. In this paper, we have proposed and investigated modified incremental Linear Discriminant Analysis (MILDA). We have compared the performance of proposed MILDA method against Pang et al ILDA in terms of classification accuracy, execution time and memory. It is found on the basis of experimental results with different face datasets that the proposed MILDA scheme is computationally efficient in terms of time and memory in comparison to batch method and Pang et al method. The experimental results also show that the classification accuracy due to MILDA, batch method and Pang et al are in complete agreement with each other.
关键词：Statistical pattern recognition; Feature extraction; Face;ecognition; Linear Discriminant Analysis
