Performance Evaluation of a Method for Isosurface Generation from Volumetric Data Sampled with a Face-centered Cubic Lattice

Abstract

In this paper, we evaluate the performance of the method for isosurface generation from volumetric data sampled with a face-centered cubic lattice. First, we made a comparison among FCC method and its modified method and MC method in terms of the error between a metaball surface and the resulting surface. Thus, it was shown that the modified FCC method decreases the dispersion of the error by smoothing vertex locations. Second, we applied these two FCC methods to medical CT data, and compared the error of resulting surfaces, the computational costs, the appearances of the surfaces with the MC method. As a result, the modified FCC method generated smoother isosurfaces in less cost with almost the same error as the MC method.