Smooth morphing of point-sampled geometry

Abstract

Based on spherical parameterization, in this paper, we put forward a smooth morphing algorithm for point-sampled geometry (PSG). Source and target PSG are first parameterized onto a unit sphere, respectively. After aligning the feature point-pairs on the two unit spheres, they are merged into a single sphere and based on it, the correspondence relation is constructed. We then use Laplacian coordinate to nonlinearly interpolate the shapes and the intermediated shapes are up-sampled using the moving least squares scheme. Experiment results demonstrate that our algorithm can generate natural intermediated shapes and visually smooth morphing sequences. © 2012 Springer-Verlag.