|
Kevin Beason Global Illumination of Isosurfaces The purpose of my research is to upgrade the quality of 3D scientific visualization by applying global illumination to isosurfaces. An isosurface of a three-dimensional (3D) function is a surface in which all the points have an identical value, called the isovalue. This isosurface may capture the shape of the brain (from MRI data), or of a nerve cell (from laser microscopy) or of neutron clusters (from a computational simulation of neutron stars). Often the shape is very complex, in which case the subtle effects of realistic lighting (like soft shadows, inter-reflection, and caustics) make the 3D structure more evident to the scientific user. Unfortunately, solving the light-transport equation for these complex surfaces may take hours to complete. My work is the first to demonstrate that pre-processing allows these high-quality renderings to be produced using the video card of an ordinary home computer, and at the same speed as the low-quality "local illumination" provided by existing commercial tools for 3D visualization. Courses taken outside of major |
