High quality triangulation of implicit surfaces

We present a new high quality tessellation method for implicit surfaces in this paper. The approach can handle arbitrary implicit functions and dynamic implicit surfaces based on skeletal primitives. We first samples the implicit surface uniformly using particle fission and floating, then reconstructs a triangular mesh from the sample points using ball pivoting algorithm (BPA). Finally, we subdivide the reconstructed surface using a 1 to 4 subdivision scheme to obtain the high quality implicit surface tessellation.