Estimation of Lower Limb Joint Angles during Walking Using Extended Kalman Filtering

Abstract

In order to evaluate the performance of an extended Kalman filter (EKF) for the estimation of lower limb joint angles generated during human walking, a pilot study is conducted using commercial inertial measurement units (IMUs) to capture 3D acceleration and angular velocity data produced by the leg during level ground, stair ascent, and stair descent walking. The inertial data from three IMUs, one mounted on each lower limb segment, are input to an EKF to estimate sagittal and coronal angles of the individual limb segments and the corresponding knee and ankle joint angles. This method is evaluated against the standard method performed using optical motion capture and inverse kinematics software. Results from three subjects are reviewed, and the promising potential of this technique for realtime applications is discussed.