Facial components extraction and expression recognition in static images

This paper deals with the emotion recognition in static images. Facial feature extraction plays a very important role in recognizing a particular emotion in humans. In this paper, the facial expressions in humans .i.e., happy, anger, sad, neutral and disgust, are recognized with the help Support Vector Machine classifier. First, a static image is taken. Then, skin region is extracted from that image using Hue Saturation Value. After skin region extraction, the right eye, the left eye and the mouth part are extracted as they are the most important part for facial expression recognition. These processes are done for every images collected in the training set. Then, Support Vector Machine classifier is used to classify which image belongs to which class category by comparing the feature vectors of the trained images. This paper produces a model which predicts a set of testing images into which class categories the image belongs to, namely anger, disgust, fear, happy and neutral.