Human Face Recognition Using Superior Principal Component Analysis (SPCA)
Arjun V. Mane, Ramesh R. Manza, Karbhari V. Kale · International Journal of Computer Theory and Engineering · 2010
Principal Component Analysis (PCA) is a statistical technique used for dimension reduction and recognition, & widely used for facial feature extraction and recognition. In this paper a cluster based SPCA face recognition method has been proposed. Experiments based on ORL face database have performed to compare the recognition rate between tradition PCA, Advanced principal component analysis (APCA), & SPCA. It is found that SPCA is giving the best classification result. has been considered, not taking the difference between the classes into account. Therefore the differences in the face images of the same person are also increasing when the differences of all images are increasing. It is disillusionary defect of PCA. This paper employed a new feature projection approach based on Advanced PCA method, doing the optimum transformation for the differences between the classes. In section 2 the Traditional PCA & Advanced PCA methodology is discussed. The Proposed methodology is discussed in section 3 and experimental results are listed in section 4. Finally, sections 5 conclude and suggest the future scope..