| 摘要 |
PROBLEM TO BE SOLVED: To provide a technology for quickly and sequentially generating avatars in a head, without viewpoint limitation in a virtual space, from images which are imaged up by plural cameras surrounding a person. SOLUTION: This method for generating the avatar of the object, on the basis of the images picked up by the plural cameras surrounding the object has a three-dimensional model acquisition step for previously acquiring a three- dimensional model with a texture of the object; a spatial coordinate value acquisition step for measuring the space position of the reference point of the object to acquire a spatial coordinate values of the three-dimensional model; an update area image acquisition step for executing both a conversion step for converting the space coordinate value of the three-dimensional model into two-dimensional coordinate values, corresponding to the images picked up by the cameras and a cutout step for cutting out update area images from the images of the object which are imaged by the cameras through the use of the two-dimensional coordinate value; and generation steps for generating the avatar, by mapping the update area mages onto the three-dimensional model. |
