EFFICIENT REPRESENTATION OF OBJECT SHAPE FOR SILHOUETTE INTERSECTION

A new shape representation the radial intersection set (RIS) is presen ted. The RIS is an object-centred model in which 2-D and 3-D boundarie s are represented via their intersection with radial lines from some s pecific origin. The RIS representation allows efficient 3-D reconstruc tion using silhouette intersection from an arbitrary number of 2-D per spective views. The relationship between the visual hull (Laurentini, 1994) of the 3-D object and the silhouettes it may generate is defined as a set intersection operator. This operator allows the direct gener ation of 3-D RIS models from silhouettes. The RIS method is shown to c ompare favourably, in terms of both speed and storage, with existing o ctree techniques. Examples of images rendered from RIS models are pres ented. These show high visual fidelity.