Facial Expression Recognition Based on Fusion Features of LBP and Gabor with LDA
Gang Bai, Wanhong Jia, Yang Jin · 2009
In this paper, the novel facial expression recognition approach using fusion features of LBP and Gabor (FFLG) is proposed. At first, local binary patterns (LBPs) and Gabor wavelet features are extracted from facial images, and then Weighed LBP Gabor complex features (WLGCF) are obtained by weighing and combining. Finally, the WLGCF as recognition features are generated with LDA algorithm to represent facial expression effectively. The experiments show our approach feasible.