Real-time Accurate Monitoring of Ten Finger Joint Angles Using a Fiber Bragg Grating Sensor-based Glove for use in Virtual Rehabilitation

Virtual reality-based exercises being immersive and interactive have become a useful tool in the rehabilitation of stroke survivors. Input systems such as instrumented gloves used in VR systems must be accurate, highly repeatable, and have very low latency to make the VR environment immersive. This paper describes an FBG sensor-based glove that offers very high accuracy (error = 0.80 degrees) and excellent repeatability (standard deviation = 0.95 degrees) in measuring the ten finger joint angles of the human hand simultaneously. The accuracy and repeatability of six of these sensors attached on the glove were tested using a pre-calibrated inertial measurement unit sensor.