| 摘要 |
PURPOSE: A face recognition method is provided to improve the feature point matching performance of SIFT algorithm. CONSTITUTION: The number of feature points is reduced by executing an algorithm of combining PCA(Principal Component Analysis) with LDA(Linear Discriminant Analysis)(S106). A feature point descriptor corresponding to the feature point is generated(S108). The generated feature point descriptor and a feature point descriptor of a face image stored in a database are matched(S110). A feature point descriptor is generated by executing the first and second processes of an SIFT algorithm(S112). |
