An Assessment Method For Upper Limb Rehabilitation Training Using Kinect

This work proposes a rehabilitation training assessment method combined with virtual reality technology (VR) to improve the training effect of stroke patients' upper limb rehabilitation. During the training, patient's upper limb location information and joint angle features are collected by Kinect2.0 somatosensory equipment. The joint angle data is filtered by the median filtering algorithm. Then a modified dynamic time warping (DTW) algorithm is used to recognize the patient's motion. Finally, the system considers both the influence of time duration and amplitude in the result of DTW algorithm to determine the training score. The experiment results show that this method can assess the user's trainig movement objectively in order to provide correct guidance.