Real time kinect based robotic arm manipulation with five degree of freedom

Abstract

We propose a real time, gesture based robotic arm manipulation using kinect sensor. This method uses kinect depth data, skeletal data and joint orientation data for end effector movement including roll, pitch and pinch. For the initial data, we extracted skeletal of the person, hand position and hand state. In closed hand state condition, the position of the hand is extracted and the three dimension data is used for inverse kinematics of the joints of the robotic arm. In open hand state, tilting and rolling of hand is used for pitch and roll respectively. In lasso state, pinch of end effector is perceived. The robotic arm used for the experimentation is Scorbot ER III. The system shows promising results for non-invasive robotic arm manipulation.