Application of the 2D-FrFT combined with fuzzy fusion classification algorithm in human emotional state recognition

In this paper we perform feature extraction by utilizing the Two Dimensions Fractional Fourier Transform (2D-FrFT). PCA is used to reduce the high-dimensional feature information and we perform FLDA on the total samples, then emotional sates are recognized based on multi-classifier fusion with fuzzy integral in decision layer. Simulations based on the Ryerson and Jaffe facial expression database show that the novel method is invalid and superior. Compared to the classical algorithm, the proposed method may flexibly extract the emotional features and effectively classify the different emotional states, so highly improve the recognition rate.