A Framework for Representing and Reasoning about Three-Dimensional Objects for Visione