Analysis of static and dynamic motion accuracy for kinect-virtual Sensei system

• Published in: ARPN Journal of Engineering and Applied Sciences
• Main Author: 2-s2.0-84943419053
• Format: Article
• Language: English
• Published: Asian Research Publishing Network 2015
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-84943419053&partnerID=40&md5=860ae1654316b60892abf07265dd67bc

Motion analysis has been widely adapted in research pertaining to biomechanics and used for many important applications such as injury reduction, sports performance enhancement and rehabilitation. Nevertheless, current available motion capture system such as the use of infrared cameras is very expensive. Microsoft Kinect has the potential to be used as an alternative low-cost motion analysis tool. Nevertheless, the standard procedure for measuring its accuracy and reliability has not been well established. Therefore, this study for the first time attempts to develop a standard procedure to assess and visualise the accuracy and repeatability of Microsoft Kinect. A single-camera system is used to capture static and dynamic motions of healthy volunteers. Adapting numerical and statistical tools, the data are analysed for the i) static motion capture (standing still with lateral hand lift) and ii) dynamic motion capture (simple lower arm movement), which are tracked by the sensor operated using open source Virtual Sensei Lite program. The variance and error value are then analysed to determine the accuracy of measurement. The study able to demonstrate average errors of less than 2% (static) and 5% (dynamic) accuracy respectively. The good results prove that the current study is important and could contribute a significant knowledge for further research in improving Microsoft Kinect functions and applications for motion analysis. © 2006-2015 Asian Research Publishing Network (ARPN).