An order of magnitude faster isosurface rendering in software on aPC than using dedicated, general purpose rendering hardware

摘要

The purpose of this work is to compare the speed of isosurfacenrendering in software with that using dedicated hardware. Input datanconsist of 10 different objects from various parts of the body andnvarious modalities (CT, MR, and MRA) with a variety of surface sizes (upnto 1 million voxels/2 million triangles) and shapes. The softwarenrendering technique consists of a particular method of voxel-basednsurface rendering, called shell rendering. The hardware method isnOpenGL-based and uses the surfaces constructed from our implementationnof the Marching Cubes algorithm. The hardware environment consists of anvariety of platforms, including a Sun Ultra I with a Creator3D graphicsncard and a Silicon Graphics Reality Engine II, both with polygonnrendering hardware, and a 300 MHz Pentium PC. The results indicate thatnthe software method (shell rendering) was 18 to 31 times faster than anynhardware rendering methods. This work demonstrates that a softwarenimplementation of a particular rendering algorithm (shell rendering) cannoutperform dedicated hardware. We conclude that, for medical surfacenvisualization, expensive dedicated hardware engines are not required.nMore importantly, available software algorithms (shell rendering) on an300 MHz Pentium PC outperform the speed of rendering via hardwarenengines by a factor of 18 to 31