Complete Local Binary Pattern for Representation of Facial Expression Based on Curvelet Transform

In this paper, we proposed a technique for facial expression representation based on combination of Curvelet Transform and Complete Local Binary Pattern (CLBP). The curvelet transform offers improved directional capability, better ability to represent edges and other singularities along curves as compared to other traditional multiscale transforms. Hence, we transform original face images to frequency domain at a specific scale and orientation using curvelet transform. Rotation invariant uniform binary patterns are extracted from the approximate sub-band using CLBP method to represent facial expressions, which form the feature vector. The efficacy of the proposed method for facial expression representation is evaluated based on expression recognition carried out using a benchmark database such as JAFFE. The recognition is performed using a Chi-square distance measure with a nearest neighbor classifier. The experimental results show that our method outperforms LBP based approach and curvelet based LBP approach.