期刊名称：International Journal of Advanced Computer Science and Applications(IJACSA)
印刷版ISSN：2158-107X
电子版ISSN：2156-5570
出版年度：2017
卷号：8
期号：1
DOI：10.14569/IJACSA.2017.080154
出版社：Science and Information Society (SAI)
摘要：The face recognition applications are widely used in different fields like security and computer vision. The recognition process should be done in real time to take fast decisions. Princi-ple Component Analysis (PCA) considered as feature extraction technique and is widely used in facial recognition applications by projecting images in new face space. PCA can reduce the dimensionality of the image. However, PCA consumes a lot of processing time due to its high intensive computation nature. Hence, this paper proposes two different parallel architectures to accelerate training and testing phases of PCA algorithm by exploiting the benefits of distributed memory architecture. The experimental results show that the proposed architectures achieve linear speed-up and system scalability on different data sizes from the Facial Recognition Technology (FERET) database.
