摘要
步态分析有助于对神经疾病患者的康复功能状态进行评价,帮助制定康复治疗方案及评定康复疗效,步态数据的采集是步态分析的前提与保障。针对现有的步态采集系统采集到的步态周期较少的问题,我们提出了一种基于多台Kinect设备,用于步态数据采集的新方法,成功延长了采集的距离。系统以建立无线局域网为基础,实现了同时控制多台Kinect;将人体模型作为采集目标,结合加权整体最小二乘法标定系统的坐标系。最后采集步态数据,实现数据的融合,取得了较好的融合效果,将步态数据与运动捕捉仪数据进行对比,证明了此系统的可行性。
Gait analysis can help to evaluate the functional status of patients with neurological diseases and develop rehabilitation programs. Collection of gait database is premise and basis for gait analysis. Many gait acquisition systems can not collect more gait cycle. Aiming at this problem,a new method of gait acquisition with Multi- Kinec was proposed in this paper and proved to be correct.The system controlled more than one Kinect simultaneously based on LAN. The total least squares algorithm and a human model as the target was used in calibrating coordinate of the system. Experimental data and good results were obtained. Finally the data of the system is compared with the 3D motion capture,which proves that this system is feasible.
出处
《生物医学工程研究》
北大核心
2016年第2期87-92,共6页
Journal Of Biomedical Engineering Research
基金
国家自然科学基金资助项目(51275282)
教育部博士点基金资助项目(20123108110009)