摘要 |
An image processing method is disclosed for generating 3-D graphical model data representing the appearance of a number of points in a 3-D region from two or more 2-D images of the region, the method comprising the steps of a) calculating, for each point, at least one variance value representing the variation, between different ones of the 2-D images in the region of the position on the 2-D images at which the point would appear if it were visible in the images; b) calculating, for each point and using a predetermined rule, a merit value from the variance values, the merit value representing the probability that, if the point lay on a visible surface in the 3-D region, the calculated variance values would be obtained; and c) assigning appearance data to each point in accordance with the merit values and the appearances of the positions on the 2-D images where the point would appear if it were visible thereon. The model is to be rendered to produce images for graphics applications.
|