Accurate Estimation of Gait Altitude Using One Wearable IMU Sensor

Abstract

We present a novel method of estimating gait altitude by using the acceleration and angular velocity data obtained from a single Inertial Measurement Unit (IMU) sensor. High-precision foot tracking and vertical positioning were achieved by using this method. The inherent drifts of IMU sensors, which may cause cumulative errors for position estimation, were reduced by using our algorithm. Firstly, a multitude-threshold detection method was used to determine the stance phase and swing phase in gait movement. Secondly, a zero-velocity update was performed in the stance phase to minimize the drift error when the single foot is stationary. Finally, in the swing phase, the motion direction was represented by quaternion, and the gait altitude was estimated using a method that combines a Complementary Filter (CF) algorithm and a Target-error Compensation algorithm. Experimental results show that this method can effectively reduce inherent drifts of wearable IMU sensors and ensure accurate estimation of gait altitude.