A Simple and Robust Stride Length Estimation Method using Foot-mounted Micro Gyroscopes

Accurate estimation of walking distance plays an important role in numerous applications, such as gait analysis, sport training and particularly pedestrian localization in indoor environment [1]. One possible pervasive solution is to make use of motion dynamics contained in the inertial signals such as frequency, maximum/minimum amplitude, to aggregate individual stride. However, the main challenge for stride aggregate is how to accurately determine the stride length; therefore, a simple and robust stride length estimation method by exploring foot-mounted gyroscope measurements is proposed in this poster. Since the stride length is only related to the leg swing forward movement, key gait events are extracted to detect the swing phase of a gait cycle from the gyroscope signal. A simple linear model between the stride length and average angular swing speed is then constructed. To calibrate the linear model for different subjects, an efficient linear model parameters estimation method is also presented. The performance of the stride length estimator has been evaluated by laboratory experiments involving 10 healthy subjects. The good preliminary results have illustrated the effectiveness of the proposed stride estimation method.