Automatic Facial Expression Recognition Based on Local Binary Patterns of Local Areas

Automatic facial expression recognition is one of the great challenges in facial expression recognition field. An algrithom of automatic facial expression recognition is proposed based on Local Binary Patterns of local areas (LLBP)in this work. First, the position of eye balls is fixed by projection method. Then the local areas of the eyes and mouthpsilas neighbourhood could be determined through the pri-knowledge of face structure. The LBP feature on the local areas is then computed as the facial feature for facial expression recognition. Finally, the recognition experiment is conducted on the JAFFE facial database, which showd the realibility of the method proposed.