Mesh Voxelization


Haris Widjaya

Discrete domain representation of 3D objects are common place in medical imaging, and simulations. The challenge of displaying 3D sampled data obtained from CT scan and MRI have largely been overcome. To interact with this object representation one requires an equivalent discrete representation of the object. This project tries to address that by converting existing library of 3D mesh models into discrete version. The current prevailing representation for 3 Dimensional object in the computer are surface meshes. The surface mesh representation have been successfully used in computer graphics due to their low storage requirements, their relative ease of rendering and manipulation. However for some application where the volume between the boundaries of the object matters, using a surface mesh to approximate the volume becomes an issue. With the abundance of available surface mesh representation of 3D objects we simply can't ignore this available library of objects. This project tries to bridge the gap between surface mesh and volumetric representation by creating a program that automatically converts surface meshes into volume data. Some of the constraints that we have to deal with are : 1. Separability condition of the volume representation (No holes on the surface of the volume object). 2. Minimality condition (We don't want to set any voxels that aren't necessary to maintain separability). Working under this constraints the program will generate the optimum representation for a particular mesh representation. We aim to further improve the state of the art voxelization through algorithm optimization, efficient lattice structure, and signal processing.