Voxelization of free-form solids represented by catmull-clark subdivision surfaces

A voxelization technique and its applications for objects with arbitrary topology are presented. It converts a free-form object from its continuous geometric representation into a set of voxels that best approximates the geometry of the object. Unlike traditional 3D scan-conversion based methods, our voxelization method is performed by recursively subdividing the 2D parameter space and sampling 3D points from selected 2D parameter space points. Moreover, our voxelization of 3D closed objects is guaranteed to be leak-free when a 3D flooding operation is performed, This is ensured by proving that our voxelization results satisfy the properties of separability, accuracy and minimality.