Novel approach of gait analysis based on wearable sensor system

摘要

In this study,a new method of anthropometric data estimation was developed for measurement of human lower limb,including hip,knee,and ankle joints based on a wearable sensor system.Combining with Kalman filter,a joint angle measurement method was used to correct outputs of the wearable sensor system which are configured as integration of triaxial accelerometers,gyroscopes,and magnetic sensors.In order to validate the accuracy of developed method,the wearable sensor system was attached to two experimental subjects,respectively,recording the change of lower limb joint angles of subjects during gait cycle.At meanwhile,a motion capture system worked as the reference system.Then,1 minute walking measurement was performed on a treadmill at three different kinds of speed.The accuracy of the joint angle measurement by wearable sensor system with Kalman filter is validated against results obtained from the reference motion capture system.The results measured by wearable sensor system showed joint angle changes were similar to those shown in camera system.