Gabor T. Herman
Graduate Center, City University of New York
Department of Computer Science
A three-dimensional (3D) object can be represented as a linear combination of "blobs" (these are spherically-symmetric smooth functions) each attached to one of an efficiently-arranged grid of points in space. The representation problem becomes that of calculating how much weight to give to each of the individual blobs based on some given data. Surfaces defined by such a representation are inherently smooth, allowing superior 3D displays based on such representations to those obtained from the traditional voxel-based representations. Volume visualization can be achieved by efficient footprint algorithms and associated hardware mappings.