EDBL - algorithm for detection and analysis of emotion using body language

This paper proposes a system which will detect emotion and mental state of a person by detecting the pose of a person. Detection of emotion will be based on body parts not on facial expressions. Work done on emotion detection is basically done on facial expression, according to the psychological study it is found that every body part shows an expression. We already have frameworks for facial expression detection but for body gestures we dont have any framework. Here we study the normal body pose first and then we can estimate the difference between a person who is calm and the person who is showing deviation. Along with the shoulder movement, we studied the hand gestures so that we can get much more promising and sound results. By combining both the results for hand and shoulder together we will get the approximate picture of the persons state of mind. Our approach is based on bodily gestures, and we are trying to detect emotions from them. Thus, to detect bodily expressions we have written an algorithm which will help us to predict the behavior of a person. The experimental result shows that by using EDBL algorithm we can infer emotions and state of mind from human pose, in terms of body gesture including shoulder and hand. EDBL algorithm helps in detecting emotions even if we exclude the facial expressions.