(a) e() Doo-Sabin surfaces. 4 Conclusion Both EDSes and quadratic NURSSes are generalizations of References 8 Z. Huang et al. /Journal of Information & Computational Science 7: 1 (2010) 1–6 (a) (b) (c) (d) Fig. 6: Doughnut model: (a) initial control mesh, (b) uniform biquadratic B-spline surface, (c) bi￾quadratic NURBS surface with a crease, (d) EDS with a dart. Doo-Sabin surfaces. 4 Conclusion This paper presents EDSes which are a modification of quadratic NURSSes [3]. EDSes retain the refinement rules for quadratic NURSSes for regular faces whereas use the refinement rules for Doo-Sabin surfaces for irregular faces. Both EDSes and quadratic NURSSes are generalizations of non-uniform biquadratic B-spline surfaces and Doo-Sabin surfaces. In comparison to quadratic NURSSes, if all the knot intervals are greater than zero, EDSes converge at extraordinary points of arbitrary valence while quadratic NURSSes may diverge for valences larger than 12. And closed-form limit point and limit normal rules are available for EDSes as well. In future work we hope to derive a parametrization for exact and efficient evaluation of EDSes following the method of Stam [13]. References [1] Doo, D., Sabin, M.: Analysis of the behaviour of recursive division surfaces near extraordinary points. Computer-Aided Design 10 (1978) 356–360. [2] Catmull, E., Clark, J.: Recursively generated B-spline surfaces on arbitrary topological meshes. Computer-Aided Design 10 (1978) 350–355