An energy preserving upscaling technique for enhanced volume rendering of medical data

Abstract

In this paper we describe an edge-directed optimization-based method for volumetric data supersampling. Our method faces the problem of partial volume effect by upscaling the volumetric data, subdividing voxels in smaller parts and performing an optimization step keeping constant the energy of each original subdivided voxel while enhancing edge continuity. Experimental tests show the good quality of the results obtained with our approach. Furthermore, we show how offline 3D upscaling of volumes can be coupled with recent techniques to perform high quality volume rendering of large datsets, obtaining a better inspection of medical volumetric data.