bebi 21: e1

Research Article

Human Motion Enhancement via Joint Optimization of Kinematic and Anthropometric Constraints

Download21 downloads
  • @ARTICLE{10.4108/eai.9-4-2021.169181,
        author={Le Zhou and Nate Lannan and Guoliang Fan and Jerome Hausselle},
        title={Human Motion Enhancement via Joint Optimization of Kinematic and Anthropometric Constraints},
        journal={EAI Endorsed Transactions on Bioengineering and Bioinformatics: Online First},
        volume={},
        number={},
        publisher={EAI},
        journal_a={BEBI},
        year={2021},
        month={4},
        keywords={D-Mocap, Mocap, Depth sensor, Kalman filters, extended Kalman filters, differential evolution, joint positions, bone length, joint angles, gait analysis},
        doi={10.4108/eai.9-4-2021.169181}
    }
    
  • Le Zhou
    Nate Lannan
    Guoliang Fan
    Jerome Hausselle
    Year: 2021
    Human Motion Enhancement via Joint Optimization of Kinematic and Anthropometric Constraints
    BEBI
    EAI
    DOI: 10.4108/eai.9-4-2021.169181
Le Zhou1, Nate Lannan1, Guoliang Fan1,*, Jerome Hausselle2
  • 1: School of Electrical and Computer Engineering, Oklahoma State University, Stillwater, Oklahoma 74078, USA
  • 2: School of Mechanical and Aerospace Engineering, Oklahoma State University, Stillwater, Oklahoma 74078, USA
*Contact email: guoliang.fan@okstate.edu

Abstract

This paper proposes a novel approach to improve the quality of human motion data captured by a depth sensor. Depth-based motion capture (D-Mocap) data suffer significant errors due to noise, self-occlusion, interference, and other technical limitations. We aim to improve 3D joint trajectories to be more kinematically admissible and to sustain the skeleton structure more anthropometrically stable. Our research integrates nonlinear Kalman filters (KF) with Differential Evolution (DE) algorithms together to take advantage of the kinematic and anthropometric constraints respectively. Specifically, we compare three nonlinear KFs in terms of their effectiveness and accuracy, including the extended Kalman filter (EKF), the unscented Kalman filter (UKF), and the Tobit Kalman filter (TKF). Both simulated and real-world D-Mocap data are used to examine the proposed algorithm. We also show the improved accuracy of six lower-body joint angles that are often used for clinical gait assessment.