Sit to stand sensing using wearable IMUs based on adaptive Neuro Fuzzy and Kalman Filter

This paper present a method for measuring the posture of a human body during different phases of sit to stand motion using inertial sensors. The proposed method fuses data from inertial sensors placed in trunk and thigh using Adaptive Neuro-Fuzzy Inference System (ANFIS) followed by a Kalman Filter (KF). The ANFIS attempts to estimate the position of shoulder of the human, at each sampling instant when measurement update step is carried out. The Kalman filter supervises the performance of the ANFIS with the aim of reducing the mismatch between the estimated and actual. The performance of the method is verified by measurements from VICON (motion analysis system). The obtained results show the effectiveness of the proposed algorithm in prediction the human shoulder position with root mean square error 0.018 m and 0.016 m in the x and y direction, respectively.