Two-Dimensional Principal Component Analysis and Concurrent Self-Organizing Maps for Face Classification

2010 
Face recognition is one of the most widely used methods for recognition of individuals. Face recognition can classify a person through a non-intrusive process where the cooperation of the person is not necessary to capture their face. The proposed algorithm for face classification is broken down into three steps. In the first step the feature matrices are obtained using the Two-Dimensional Discrete Cosine Transform (2D-DCT) and Two-Dimensional Principal Component Analysis (2DPCA). The training of Concurrent Self-Organizing Maps (CSOM) is achieved in the second step by using the face's feature matrix. And finally, the third step extracts the feature matrix from the query image and classifies it using the CSOM network. To verify the efficiency of the algorithm, the tests have been done using the "The ORL Database of Faces" provided by the AT&T Laboratory from Cambridge University. The performance of the algorithm was satisfactory in relation to other proposed algorithms for face recognition found in the literature.
    • Correction
    • Cite
    • Save
    • Machine Reading By IdeaReader
    10
    References
    1
    Citations
    NaN
    KQI
    []