抄録
In the recent years, motion capture systems are used in order that robots may achieve human motion. Most of those systems need attached markers on human body and several sensors. So makerless motion capture system is needed. In this paper, we propose a posture estimation method using monocular camera. This method detects each body part (arms, legs, head, torso and so on) without markers using motion feature and shape feature. A posture is estimated by constructing the detected parts based on human body structure model. A markerless motion capture system is realized by using this method. In the experiment, we estimated human arm posture using this method.
