Estimate Hand–Finger Position With One Magnetometer and Known Relative Orientation
IEEE Transactions on Instrumentation and Measurement
Inertial-sensor-based hand motion tracking has become a well-accepted method in many clinical applications, including rehabilitation of the upper extremity. However, major drawbacks are that a sensor on each segment and kinematic chain rule is required. Thus, errors accumulate through the kinematic chain rule. Also, the length of each segment needs to be measured for each user. We propose a novel method in which a permanent magnet on the dorsal side of the hand combined with an inertial and
... etic measurement unit (IMMU) on the fingertip is used to estimate the position of the thumb and index fingertip relative to the position of the dorsal side of the hand. The biggest advantages of the proposed method are that no prior information and kinematic chain rule are needed. Besides, the required number of sensors is low. A Levenberg-Marquardt (L-M) approach is presented to estimate the relative positions with the known orientations. The performance is demonstrated in multiple experiments in which various movement tasks were performed. The most complex task in which participants performed reaching and grasping movements based on the action research arm test (ARAT) resulted in a median distance error between thumb and index finger of 9.6%.