Extraction of curvelet based RLBP features for representation of facial expression

In this paper, we proposed a technique for representation of facial expression based on extraction of Robust Local Binary Pattern (RLBP) features in curvelet domain. The curvelet transform exhibit improved directional capability, better ability to represent edges and other singularities along curves as compared to traditional multiscale transform such as wavelet transform. Hence, we transform original face images to frequency domain at a specific scale and orientation using curvelet transform. Noise and illumination invariant features are extracted from approximate sub-band using robust local binary pattern, which forms the feature vector of facial expression. The proposed method is evaluated based on facial expression recognition carried out using a benchmark database such as JAFFE. The facial expression recognition is performed using a chi-square distance measure with a nearest neighbor classifier. Experimental results show that our approach outperforms other popular LBP based approaches.