| 摘要 |
<p>PROBLEM TO BE SOLVED: To provide a device and a method for inputting action, which recognizes the action of a user with a little markers or sensors and makes avatar in a virtual space act almost as the user acts. SOLUTION: First of all, the three-dimensional (3D) positions and directions of the head, wrists and a central point are detected by magnetic sensors attached to the head, both the wrists and the back of the user. Next, calibration is performed by spreading both arms, the directions of sensors are matched and the size of the avatar is calculated from the position of the magnetic sensor on the head. Next, avatar information showing the calculated size and the detected position and direction is transmitted to the image display terminal equipment of the other user. In each of image display terminal equipment, the positions and direction of the shoulders of the avatar are calculated based on the avatar information and the moving positions and directions of respective joints of the arms are predicted from the positions and directions of the wrists while using the numerical calculation method of inverse kinematics. The virtual space making the avatar act, is generated from this calculated result and the avatar information and a 3D virtual space is displayed on a display.</p> |
