Feature extraction method based on the generalised Fisher discriminant criterion and facial recognition

In this paper we view the optimal set of discriminant vectors as a global t ransform, and consider its separability from a global view point. Based on this idea, the concept of a generalised Fisher discriminant criterion and t hat of a generalised optimal set of discriminant vectors is introduced. Alt er that, a new algorithm is given to calculate the generalised optimal set of discriminant vectors defined in this paper, which is particularly suited to the case of a small number of samples where the scatter matrix is singu lar. It is then applied to an experiment on human facial recognition, and t he results show that the new algorithm is superior to existing methods in t erms of correct classification rate.