摘要 |
PURPOSE:To obtain the real-time recognizing and composing method for the human image which can reproduces the motion of a man in real time with a human model by putting the motion detection part of the man and the motion composition part of the human model together. CONSTITUTION:A three-dimensional model generation part 20 generates the three-dimensional model for the man 2 in a real space 1 and a real-time detection part 30 detects the expression, the rotation of the head, the motion of fingers, and the rotation of the human body as the motion of the man in a communication conference. On the basis of the detected motion information, a real-time generation part 40 generates the human image and a virtual space 44 is displayed as a virtual space 4 including the human image 3 on a reception- side stereoscopic display 5. Consequently, reception-side conference attendants 6 and the transmission-side man 2 have a conference as if they were present at one place. |