Comparison of Marker-Based and Marker-Less Systems for Low-Cost Human Motion Capture

Abstract

To generate graphic simulation of human motion, marker-based optical motion capture technology is widely used because of the accuracy and reliability of motion data provided by this technology. However, tracking of human motion without markers is very desirable on the factory floor because the human operator does not need to wear a special suit mounted with markers and there is no physical interference with the manufacturing or assembly operation during the motion tracking. In this paper, we compare marker-based and marker-less motion capture systems. First, the operational principles of these two different types of systems are compared. Then the quality of motion data obtained by a marker-less system using Kinect is compared with that obtained by a marker-based system using Optitrack cameras. The comparison also includes the accuracy of body joint angles and variations in body segment lengths measured by the two different systems. Furthermore, we compare the human motion simulation developed in the Jack digital human modeling software using the data captured by these two systems when a person is performing a fastening operation on a physical mockup of the belly section of an aircraft fuselage.