抄録
Today the motion capture system is expected which can capture the human motion with online and on-board system and no markers. Such system usually consists of the color camera and has the difficulty of dividing the subject from the background. In this paper, we present the method of detecting the feature point of the human from the depth image in order to develop a markerless motion capture system by using one depth image camera and the computer vision processing method. This system validates the human recognition without the color information from the color cameras and validates the human motion capture based on the knowledge of the human figure.
