Bounding spheres are utilized frequently in many computer graphics and visualization applications; and it is not unusual that the computation of the spheres has to be done during run-time at real-time rates. In this paper; an attractive algorithm for computing bounding spheres under such conditions is proposed. The method is based on selecting a set of k extremal points along s = k/2 input directions. In general; the method is able to compute better fitting spheres than Ritter’s algorithm at roughly the same speed. Furthermore; the algorithm computes almost optimal spheres significantly faster than the best known smallest enclosing ball methods. Experimental evidence is provided which illustrates the qualities of the approach as compared to five other competing methods. Also; the experimental result gives insight into how the parameter s affects the tightness of fit and computation speed.
CR Categories: F.2.2 [Analysis of Algorithms and Problem Complexity]: Nonnumerical Algorithms and Problems—Geometrical problems and computations; I.3.5 [Computer Graphics]: Computational Geometry and Object Modeling
Keywords: Bounding sphere; enclosing ball; extremal points; computational geometry; computer graphics