A Space-Filler Mechanism to Recognize Body Language of Hands

Among all the species, human hands become outstanding and unique in performance. What is special about human hand is not what it can accomplish but what it can communicate to outside. Hand pose can say a lot about a person's psychological state. Hands are expressive so that they can be used to tell stories, communicate with sign language and display innermost ideas in one's mind. Hence the hand can often be used for nonverbal communication. Through this paper we attempted to drive a mechanism to perceive hand posture using the depth information received from upper arm and fore arm separately. We used positioning of upper and fore arms within the activity space of a person as to determine its relative pose with respect to the person's torso. As dimensions of human body change from person to person, we considered all the measurements with respect to that particular individual. We implemented this method upon a service robotic platform for evaluation.