A face identification algorithm using support vector machine based on binary two dimensional principal component analysis

The two dimensional human face image (2DHFI) matrices have to be previously transformed into one dimensional image vectors row by row or column by column In the human face recognition schemes based on the one dimensional principal component analysis (1DPCA), such that the 1DPCA scheme is difficult in accurately evaluating the human face image covariance matrix and is time-consuming in determining the eigenvectors. The two dimensional principal analysis (2DPCA) schemes evaluate the HFI covariance matrix more accurately and determine the corresponding eigenvectors more efficiently than 1DPCA schemes. But, the 2DPCA schemes need many more coefficients for HFI representation than 1DPCA schemes. The binary principal component analysis (B-PCA) replaces floating-point multiplications with integer additions to significantly reduce the time consumption of the procedure. This paper utilizes the binary two dimensional principal component analysis (B2DPCA) to construct an effective human face identification system. The presented algorithm combines the scaling process, histogram equalization process, binary two dimensional principal component analysis (B2DPCA) process, and support vector machine (SVM) scheme to construct a human face identification system. The experimental results show that the presented algorithm has good efficiency for human face identification.