International Research Journal of Education and Technology

The universally common mode of interaction is the human emotions. Thus, there are several advantages of automated recognition of human facial expressions. The primary objective of the proposed framework in this paper, is to classify a person’s facial expression into anger, contempt, disgust, fear, happiness, sadness and surprise. Firstly, CLAHE is performed on the image and the faces are identified using a histogram of oriented gradients. Then, using a model trained with the iBUG 300-W dataset the facial landmarks are predicted. Using the proposed method with the normalized landmarks, a feature vector is calculated. With this calculated feature vector, the emotions can be recognized using a Support Vector Classifier. The Support Vector Classifier was trained and tested for system accuracy using the famous Extended Cohn-Kanade database.