Point-based Volume Grpahics

In general, volume graphics is the subfield of computer graphics that employs a volume buffer for scene representation and is concerned with synthesizing, manipulating, and rendering such scenes. In this context we refer to grid-based volume graphics, when the volume data has a grid-based representation. To overcome the limitations and problems, especially with large data, implied by traditional grid-based volume graphics, we propose point-based volume graphics. A paradigm shift from grids, as a spatial-oriented representation, to points, as an object-oriented representation, promises considerable advantages. Removing the grid-structure and introducing a point-based representation of volumetric data has the potential to simplify, and often even completely void many problems incurred by a grid-based representation. The main purpose of this project is to address the major issues which arise when performing a paradigm shift from grid-based to point-based volume graphics. At first we will focus on Pointization. This is a process which is concerned with converting analytical volume data or discrete volume data into a set of point primitives that best represents the underlying volumetric function. Instead of converting grid-based volumetric data into a point-based representation, the raw data provided by scanning devices, such as CT or MR scanners, is directly converted (pointized) into a point-based representation in order to avoid inaccuracies resulting from reconstructing of data sampled onto a grid. In a next step, in order to prove the usability of our point-based representations we will develop approaches to support typical applications such as rendering and segmentation. Finally all these approaches are mapped onto the graphics hardware and integrated into one powerful full-blown point-based volume graphics application. Once the application is fully functional all the approaches will be tested with extremely large data sets.

In general, volume graphics is the subfield of computer graphics that employs a volume buffer for scene representation and is concerned with synthesizing, manipulating, and rendering such scenes. In this context we refer to grid-based volume graphics, when the volume data has a grid-based representation. To overcome the limitations and problems, especially with large data, implied by traditional grid-based volume graphics, we propose point-based volume graphics. A paradigm shift from grids, as a spatial-oriented representation, to points, as an object-oriented representation, promises considerable advantages. Removing the grid-structure and introducing a point-based representation of volumetric data has the potential to simplify, and often even completely void many problems incurred by a grid-based representation. The main purpose of this project is to address the major issues which arise when performing a paradigm shift from grid-based to point-based volume graphics. At first we will focus on Pointization. This is a process which is concerned with converting analytical volume data or discrete volume data into a set of point primitives that best represents the underlying volumetric function. Instead of converting grid-based volumetric data into a point-based representation, the raw data provided by scanning devices, such as CT or MR scanners, is directly converted (pointized) into a point-based representation in order to avoid inaccuracies resulting from reconstructing of data sampled onto a grid. In a next step, in order to prove the usability of our point-based representations we will develop approaches to support typical applications such as rendering and segmentation. Finally all these approaches are mapped onto the graphics hardware and integrated into one powerful full-blown point-based volume graphics application. Once the application is fully functional all the approaches will be tested with extremely large data sets.