New face expression recognition using polar angular radial transform and principal component analysis

Abstract

This paper presents a new method for facial expression recognition (FER) using a polar mathematical development based on the angular radial transformation (ART). This method is combined by polar angular radial transform (P-ART) and principal component analysis (PCA). The new ART is a powerful descriptor in terms of robustness, description form and way more information-rich compared to the conventional Cartesian descriptor. Support vector machine (SVM) training is used to recognise the facial expression for an input face image. Finally, the experimental results show the performance of the P-ART and the PCA. The fusion of these two techniques can be better than other existing methods of recognition of facial expression. During the experiment, the basis of facial given Japanese female facial expression (JAFFE) and the Cohn-Kanade databases has been used.