基于Kinect的下肢体康复动作评估系统

摘要

针对中风患者下肢体的训练需求,设计了一个康复训练评估系统.利用Kinect 2.0体感设备实时采集下肢体康复动作的序列信息,提取相关关节的角度序列特征,与动作库中的标准动作序列特征对比后,计算相关度,再加入时间参量,综合得出评估结果.试验结果表明:该系统能准确地识别和引导受试者的康复动作;并且可以通过组建不同的姿势库来评估更多的康复动作.%Aiming at demand for training under limbs of stroke patients,design a rehabilitation training assessment system.System real-time collect sequence information of lower limb rehabilitation action utilizing Kinect 2.0 somatosensory equipment.Extract angle sequence characteristics of relevant joints,after comparing with the action library standard action sequence characteristics,calculate correlation,add time parameter,comprehensive assessment of results is obtained.The results show that this system can accurately identify and guide the participants' rehabilitation,and more rehabilitation actions can be assessed through formation of different posture library.