| 摘要 |
The invention provides a method of providing a volumetric representation of a three-dimensional object (10). A plurality of voxels (22) are defined in a three dimensional space and voxels are categorised as either foreground or background according to at least one silhouette image of the object. Foreground voxels are assigned a first binary value and background voxels a second binary value. An intermediate value is assigned to voxels, preferably according to an estimate of the ratio of background to foreground contained within said voxel, thereby providing more information than just first and second binary values. Several examples are given for how to determine the proportion, or fraction, of each voxel or portion thereof Voxels may be sub-sampled to categorise voxels as foreground or background or the volumetric representation may be provided as the intersection of generalised cones of the silhouette image with volume being removed from the representation in dependence on the geometry of the intersections. The object may also be represented over time as a four dimensional model defining a hypersurface to which a smoothing operator may be applied. |
