Discriminative common vector for face identification

In this paper, it is proposed a facial biometric identification system, using discriminative common vector. This method reduces the number of characteristics of the different images from the database and selects the most discriminative of them. In this work, transformed domains, such as discrete cosine transformed (DCT), discrete wavelets transformed (DWT), principal component analysis (PCA), linear discriminative analysis (LDA) and independent component analysis (ICA) are also used. As classifier systems a support vector machines (SVM) and a neuronal network (NN) have been utilized. With the above system, a simple and robust system with good results has been obtained. Using DCV, our experiments have reached a success rate of 99.13%plusmn0.23 for ORL and 99.4%plusmn0.35 for Yale.