Human Expression Recognition Based on Facial Features

Facial expression analysis is rapidly becoming an area of interest in computer science and human-computer interaction design communities. The most expressive way human displays emotion is through facial expressions. The contours of the mouth, eyes and eyebrows play an important role in classification of facial expressions. It can be classified into some classes like happiness, sadness, disgust, fear, anger, surprise and neutral. In our study we have used facial parts (two eyes, nose tip, mouth and eyebrow corners) and measured distances from those detected parts. We have used six features and used Canberra Distance (CD) for the recognition of facial expression. Increasing facial expression recognition rate is the main focus of our work. The results show that our system performs better than some other conventional methods.