| 摘要 |
Complex and/or deformable objects are represented, under a specified scene condition, with an improved trade-off between necessary computing power, on the one hand, and a realistic representational impression, on the other hand, by looking up a first view that shows the objects under a scene condition similar to the predefined scene condition using the predefined scene condition, in a database (16) comprising, for each of a multitude of different scene conditions, a view of an object (34) which shows the object under a respective scene condition, and associated texture map mapping information specifying a mapping between a projection of the object in the respective view and a predefined texture map corresponding to a desired appearance of the object. Subsequently, a second view is created by adapting the first view to the predefined scene condition, and on the basis of the texture map mapping information, the object is textured, so that the look of the object (34) in the second view corresponds to the predefined texture map (40, 58). |
