A Novel Method of Face Recognition Based on the Fusion of Classifiers

Rough neural network has the advantage of reducing training time and optimizing network topology architecture. According to such attribute, a novel face recognition method is presented based on multi-features using fusion of multiple rough neural network classifiers. First, three different feature domains are used for extracting features from input images, including IO (the interest operator), PCA (the principal component analysis) and FLD (the Fisher´s linear discriminant). Second, three independent rough neural network classifiers are used for recognition in three different feature domains respectively. Then a modified vote rule is used for decision-fusion of multiple face recognition classifiers. Experimental results show that the face recognition method proposed in this paper possesses good classification accuracy and the reliable recognition rate