A Simple Algorithm for Assimilating Marker-Based Motion Capture Data During Periodic Human Movement Into Models of Multi-Rigid-Body Systems

Taishin Nomura,Takuya Inoue,Yasuyuki Suzuki

doi:10.3389/fbioe.2018.00141

Taishin Nomura, Takuya Inoue + Show 1 more

Open Access

https://doi.org/10.3389/fbioe.2018.00141

Copy DOI

Abstract

Human movement analysis is often performed with a model of multi-rigid-body system, whereby reflective-marker-based motion capture data are assimilated into the model for characterizing kinematics and kinetics of the movements quantitatively. Accuracy of such analysis is limited, due to motions of the markers on the skin relative to the underlying skeletal system, referred to as the soft tissue artifact (STA). Here we propose a simple algorithm for assimilating motion capture data during periodic human movements, such as bipedal walking, into models of multi-rigid-body systems in a way that the assimilated motions are not affected by STA. The proposed algorithm assumes that STA time-profiles during periodic movements are also periodic. We then express unknown STA profiles using Fourier series, and show that the Fourier coefficients can be determined optimally based solely on the periodicity assumption for the STA and kinematic constraints requiring that any two adjacent rigid-links are connected by a rotary joint, leading to the STA-free assimilated motion that is consistent with the multi-rigid-link model. To assess the efficiency of the algorithm, we performed a numerical experiment using a dynamic model of human gait composed of seven rigid links, on which we placed STA-affected markers, and showed that the algorithm can estimate the STA accurately and retrieve the non-STA-affected true motion of the model. We also confirmed that our STA-removal processing improves accuracy of the inverse dynamics analysis, suggesting the usability of the proposed algorithm for gait analysis.

Highlights

Human movement analysis is performed in various fields, including biomechanics, physiology, orthopedics, neurology, and sports science, playing an important role for understanding physical functions, motor control, and motor dysfunctions
Seven colored curves represent θH(GA)T or θm that were estimated by the naive algorithm from each of the soft tissue artifact (STA) profiles for seven subjects shown in Figure 3, for which the colors in each panel of Figures 4A–D correspond to those in each panel of Figure 3
We proposed a simple but efficient algorithm to assimilate data of motion-captured marker positions affected by soft tissue artifact (STA) into models of multi-rigid-body systems

Summary

Introduction

Human movement analysis is performed in various fields, including biomechanics, physiology, orthopedics, neurology, and sports science, playing an important role for understanding physical functions, motor control, and motor dysfunctions (e.g., see Harris and Smith, 1996; Winter, 2009; Lu and Chang, 2012 for a review). A goal of the data assimilation process is to make the motions captured from an experimental subject physically consistent with those of the model, while they are as close as to the actual motions of the subject (Cappozzo et al, 2005). This is because, if no assimilation processing is performed, the original, non-assimilated motion data exhibit several inconsistencies with the model, such as temporal variations in the length of each link, which should not happen under the rigid-body-model assumption. For practical applications, it is preferable that the accurate data assimilation can be performed with a small number of markers used for the motion capture (Simon, 2004)

Methods

Results

Conclusion