Face Recognition with Locally Linear Embedding on Local Binary Patterns

In this paper, A new approach to face recognition is constructed by combining the local binary pattern (LBP) operator and locally linear embedding (LLE). LBP is an effective low-cost image descriptor to extract facial texture feature which represents the local structure of face images. LLE is an excellent non-linear data dimensionality reduction method. Its main optimization only involves a sparse eigenvalue problem and do not involves local minima. The new approach benefits from the advantages of both LBP and LLE. The proposed algorithm is experimented on ORL database. Extensive experiments are carried out to compare with other common methods such as LDA and LLE. The experiment results show that the combination of LBP+LLE provides better performance than that of those traditional algorithms and prove the effectiveness of the proposed algorithm.