Content area
Full text
Abstract
In recent years, affordable motion tracking technology has become increasingly pervasive and Kinect-based motion tracking has become popular in the field of exergaming. In the exergaming system, however, it is necessary to accurately measure and recognize the users' exercise postures in order to prompt users to do the exercises properly. This research focuses on the evaluation of accuracy of skeletal tracking by the Kinect sensor for user exercise posture estimation in comparison to the marker-based motion capture system. The detailed analysis of the exercise posture tracking accuracy shows that the Kinect sensor tracks the exercise postures relatively accurately within an error range.
Key Words: Kinect, Motion capture system, Exergame, Exercise Posture Recognition
1.Introduction
Recent developments in medical science and technology have resulted in the prolongation of the average life expectancy of humans and the aging society has progressed globally. As a result of the global aging population, there have been many research attempts for the development of policies to cope with the increasing elderly population and to develop various technologies for the promotion of physical and cognitive activity in the elderly. One of main streams is exergaming [1-3].
Exergaming combines exercise and video games. By making exercise into a game content, exercise become more fim and it encourages people to enjoy more exercise, which leads to a healthy life. Thanks to the popularity of Nintendo's Wii Sport, Wii Fit, and Microsoft's Xbox fitness games, the general public's interest in exergaming has increased. In particular, the Xbox fitness games track users' body movements using the Kinect sensor so that it enables users to enjoy the game without the need for additional controllers. Indeed, the positive effects of exergaming on health outcomes have been reported [4-6].
Chong Chong Step Master is an exergaming system which is designed for the improvement of balance and cognitive function in the elderly [3]. This gaming system is built on top of the Unity3D game engine. It uses the Smart mat sensor to recognize the walking motion of the user and the Kinect-based posture estimation system to recognize the upper and lower body exercise posture simultaneously. The Kinect-based posture estimation system extracts feature data from the skeleton model of Kinect SDK and distinguishes the users' exercise postures by comparing them...