Emotion detection using various geometrical position of the facial components

Facial image recognition technology has seen many developments in the recent past. In this paper, we have proposed one such intelligent system to recognize human facial expressions automatically. The system has been segmented into three phases. The first phase involves preprocessing and face detection. The second phase consists of feature points extraction i.e. for different facial components like lips, eyes, and eyebrows, are identified. Features are extracted by taking into account the attributed values of both normal and distorted facial components. The third phase involves data reduction method which reduces the dimension of feature points of the facial components and the face is classified based on the emotions it exhibit. The performance of the proposed approach is evaluated on JAFFE and found that the proposed system performs well.