Combining multi-resolution wavelets with principal component analysis for face recognition

The modern technological development in the field of communications and electronic systems has contributed to reducing the complexity of the application systems. This paper introduces an advanced face recognition technology, using multiresolution wavelet transform and principal component analysis (PCA). In this method, five levels of discrete wavelet transform (DWT) are used, where each level is subjected to one type of wavelet family. Then, all images are projected to the principal component domain to produce the training features array with further reduction in the processing data. The classifier of this method is the Euclidean distance, such that the minimum distance guides to know the index of anonymous person. The proposed method achieves 99.5% and 98.89% using the ORL and Yale datasets, respectively. Experimental results show that the proposed method outperforms other approaches that used the DWT-PCA technique.