作　　者： ; ; ; ;

机构地区： 中山医学院

出　　处： 《微型电脑应用》 2014年第3期32-34,共3页

摘　　要： 尽管大部分康复训练需要在医院进行,但是将训练适量"搬运"到患者家中并把结果定期反馈回医院以制定下一步治疗计划,不仅能减轻康复医师的工作,也可为患者提供更轻松自在的训练环境和完善的后期治疗。在Visual Studio 2010(VS2010)开发环境下,尝试利用Kinect传感器捕捉患者运动,获取原始深度数据流并从中采集患者全身20个关节点的信息,在此基础上构建一个动态实时的康复训练软件系统。该软件系统可存储历史训练细节,并依据标准动作模板生成的评分函数得到训练效果分数,在预设时间到达后展示给患者与治疗师同时输出训练记录,方便后期实现部分网络交流功能。目前本系统主要收录的动作有五个分别为:击掌、举手上伸、下垂合手、右腿划圈、左腿划圈。 Although rehabilitation training needs to be taken in hospital, moving considerable amount of training to patients＇ home and sending feedbacks to hospital regularly has lots of benefits. As it can reduce the workload of rehabilitation therapists, supply patients with a more comfortable environment and better late treatment. This thesis is aimed at discussing about the feasibility of community household rehabilitation system and introducing the preliminary implementation and improvement. The system uses Ki- nect sensor to track the motion of patients, get original depth data stream and collect the information of the 20 joints of the patient. It creates a dynamic real-time changing stickman and practicing environment on the platform of Visual Studio 2010 with the data basis before. At the meantime, the computer uses the algorithm to compare the collected data with the model, store the history of the details of the practice and show the results to therapists and patients when the designed time is reached. And it also can output the training records for later network social communication. There are five main actions and they are： clapping hands, raising hands and hold- ing ,dropping hands, right leg cycling ,left leg cycling.

关 键 词： 虚拟现实 家庭康复 传感器 动态实时火柴人 自动评分

领　　域： [自动化与计算机技术] [自动化与计算机技术]