惯性测量装置
加速度计
均方误差
方向(向量空间)
流离失所(心理学)
大地测量学
单调的工作
步态
骶骨
垂直位移
步态分析
相关系数
计算机视觉
数学
计算机科学
人工智能
工程类
统计
地质学
几何学
医学
结构工程
物理医学与康复
心理学
操作系统
古生物学
心理治疗师
物理疗法
作者
Stefano Cardarelli,Alessandro Mengarelli,Andrea Tigrini,Annachiara Strazza,Francesco Di Nardo,Sandro Fioretti,Federica Verdini
标识
DOI:10.1109/tim.2019.2962295
摘要
In this article, a self-contained procedure to estimate the vertical, medial–lateral, and anterior–posterior displacement of a single sacrum-worn inertial measurement unit (IMU) is presented, which can be related to the human body center of mass (CoM) displacement during treadmill walking through an adaptation of the sacral marker method. Furthermore, a magnetometer-free custom sensor-fusion algorithm for orientation estimation is proposed alongside a practical alignment procedure to refer relative IMU orientation estimation to a ground-fixed reference frame. Twelve healthy subjects performed two trials of treadmill walking at 3, 4, and 5 km/h for 150 s, with a sacrum-worn IMU. Orientation and displacement estimations were then compared with those obtained from an optoelectronic measurement system. Roll, pitch, and yaw angles showed root mean square error (RMSE) lower than 2° for walking trials at 3, 4, and 5 km/h, with Pearson’s correlation coefficient higher than 0.90 for each angle. Displacement accuracy was evaluated in terms of peak-to-trough distances and RMSE. Mean errors resulted lower than 1 mm for each axis of interest and for each gait speed, with RMSE not higher than 2.5 mm. The proposed off-line algorithm can be used in low-budget and infrastructure-free environments, to achieve reliable CoM displacement estimation during cyclic activities such as treadmill walking.
科研通智能强力驱动
Strongly Powered by AbleSci AI