Feature estimation for efficient streaming

Abstract

In this paper we propose a method to estimate the number of primitives that are extracted from an iso-surface extraction procedure. A decimated version of the data is used to estimate the number of triangles at the finest resolution through a combinatorial audit of the features (iso-contours) and their changes across resolutions. The estimates allow for the efficient implementation of visualization pipelines across a spectrum of hardware platforms. We demonstrate the efficacy of our methods through suitable experiments with The Visible Woman dataset.