Wearable Device to Record Hand Motions based on EMG and Visual Information

Human hands play a very important role in the interaction with the external world. The hands can realize various movements using their complex structure of skeleton, tendons and muscles. Analyzing the type, frequency and duration of the grasping motions in our daily life is important for the development of robotic hands and rehabilitation. In previous studies, the hand motion has been analyzed often in well-controlled experimental environments. In this research, we develop a wearable device which is attached to the forearm to analyze the hand motion in daily-life activities. The developed device can record the electromyogram (EMG) and joint angles of the user's hand simultaneously, without affecting the hand movements and grasping motions in daily-life activities. We use two commercially-available devices: the hand tracker Leap motion and the EMG-based sensor Myo, which is a gesture control armband. We propose a recognition method which uses the data acquired with these two sensors to recognize six representative types of grasping motions, ubiquitous in daily-life activities. In the experiments, we measured hand motions using the developed device on three subjects manipulating objects from a standard hand function assessment kit, and confirmed the effectiveness of the proposed method. The average recognition rate of all movements was 87.3%.