Human Motion Evaluations Based on Energy Consumption and Joint Angle Similarity

This paper presents an evaluation of human motion in handling tasks. A human skeleton model is built based on the joint nodes recognized with the Azure Kinect DK depth camera. Then, the energy consumption during the handling process is calculated and evaluated. To further evaluate the human motion, six volunteers are chosen to conduct the same handling task. Both energy consumptions and motion similarities are compared. Useful suggestions on improvement of the handling posture are finally given.