Research On Voxelization Algorithm Of Supporting Volumetric 3D Display And Development Of Visual Platform

Volumetric 3D display is a frontier field of computer graphics and display technology. Volumetric 3D display is a true 3D display device; it is different from traditional planar display. Using Volume visualization technology, it can provide 3D scene display with highly realistic physical depth.In this thesis, we focus our attention on the acquiring of 3D volumetric data for volumetric display system. Volume visualization describes objects using discrete voxels, but volume modeling tools are still limited to laboratory now, most of 3D model are constructed in surface modeling method. For volumetric 3D display system, the conversion from a surface representation to a volumetric mode, which called voxelization is a most simple and direct approach to obtain volumetric data. Voxelization utilizes spatial sampling to get volumetric data. The Cartesian lattice is traditionally used in sampling due to its simplicity. From a sampling theory we know that the choice of lattice structure can affect the quality of the sampled data.We have shown that the BCC lattice is a more effective sampling structure in 3D space. On the basis of the method that uses a Euclidean distance criterion to voxelize polygonal meshes, we extend a new algorithm for both of the Cartesian lattices and BCC lattices. We've voxelized 3DS model in the two lattices successfully, and a visual simulation platform is development. The experiment shows that by using an optimal BCC lattice, one can reduce the lattice resolution by 30% compared to a Cartesian lattice to get the same amount of information, and the BCC lattice captures the same surface detail information as the Cartesian lattice with 40%～50% less voxels. This is very available for saving storage space and transfers bandwidth.