Smoothness analysis of subdivision schemes by proximity (Q851117)

Curve subdivision schemes in general consist of repeated refinement of control polygons. This paper is a sequel to the work of \textit{J. Wallner} and \textit{N. Dyn} [Comput. Aided Geom. Des. 22, 593--622 (2005; Zbl 1083.65023)], which defines a wide class of curve subdivision schemes on manifolds, and analyzes convergence and \(C^1\) smoothness. The analysis of such a nonlinear scheme is performed by its proximity to the corresponding linear scheme from which it was derived. In that sense the nonlinear scheme is a certain perturbation of the linear one. The author gives upper bounds on the magnitude of the perturbation which are sufficient for \(C^k\) smoothness of the limit curves of a nonlinear scheme, if the original linear scheme has the same property. The main result of this paper is that convergent geodesic/projection/Lie group analogues of a certain class of factorizable linear schemes have \(C^2\) limit curves.