Daqing Xue

Efficient Splatting Using Modern Graphics Hardware

Abstract Interactive volume rendering for data set sizes larger than one million samples requires either dedicated hardware, such as three-dimensional texture mapping, or a sparse representation and rendering algorithm. Consumer graphics cards have seen a rapid explosion of performance and capabilities over the past few years. This paper presents a splatting algorithm for direct volume rendering that utilizes the new capabilities of vertex programs and the OpenGL imaging extensions. This paper presents three techniques: immediate mode rendering, vertex shader rendering, and point convolution rendering, to implement splatting on a PC equipped with an NVIDIA GeForce4 display card. Per-splat and per-voxel render time analysis is conducted for these techniques. The results show that vertex-shader rendering offers an order of magnitude speed-up over immediate mode rendering and that interactive volume rendering is becoming feasible on these consumer-level graphics cards for complex volumes with millions of voxels.

Rendering Implicit Flow Volumes

Traditional flow volumes construct an explicit geometrical or parametrical representation from the vector field. The geometry is updated interactively and then rendered using an unstructured volume rendering technique. Unless a detailed refinement of the flow volume is specified for the interior, information inside the underlying flow volume is lost in the linear interpolation. These disadvantages can be avoided and/or alleviated using an implicit flow model. An implicit flow is a scalar field constructed such that any point in the field is associated with a termination surface using an advection operator on the flow. We present two techniques, a slice-based three-dimensional texture mapping and an interval volume segmentation coupled with a tetrahedron projection-based renderer, to render implicit stream flows. In the first method, the implicit flow representation is loaded as a 3D texture and manipulated using a dynamic texture operation that allows the flow to be investigated interactively. In our second method, a geometric flow volume is extracted from the implicit flow using a high dimensional isocontouring or interval volume routine. This provides a very detailed flow volume or set of flow volumes that can easily change topology, while retaining accurate characteristics within the flow volume. The advantages and disadvantages of these two techniques are compared with traditional explicit flow volumes.

Volume interval segmentation and rendering

In this paper, we segment the volume into geometrically disjoint regions that can be rendered to provide a more effective and interactive volume rendering of structured and unstructured grids. Our segmentation is based upon intervals within the scalar field, producing a set of geometrically defined interval volumes. We present many advantageous properties in using interval volumes, and provide several new rendering operations or shaders to provide effective visualizations of the 3D scalar field. In particular, we demonstrate new technologies that allow interval volumes to be rendered interactively and/or used to reduce the amount of rasterization or rendering primitives in a volume renderer. We illustrate the use of interval volumes to highlight contour boundaries or material interfaces. Several surface shaders that can easily be integrated in the volume renderer are presented. To construct the interval volumes, we cast the problem one dimension higher, using a higher-dimensional isosurface construction for interactive computation or segmentation. The algorithm is independent of the dimension and topology of the polyhedral cells comprising the grid, and thus offers an excellent enhancement to the volume rendering of unstructured grids. We present examples using hexahedral and tetrahedral cells from time-varying and multi-attribute datasets.

Light propagation for mixed polygonal and volumetric data

Some applications require scenes mixing polygonal and volumetric objects and shadows make the scenes more realistic. This paper describes a shadow algorithm for mixed polygonal and volumetric data, including the generation of soft shadows for area light sources. Our volume shader leverages advanced graphics GPU for an accelerated and feasible solution. The shadow and soft shadow algorithm applies to all combinations of volumes and polygons, without any restriction on the geometric positioning and overlap of the volumes and polygons. For realistic rendering where we have a high albedo participating media, multiple scattering is significant. We extend our algorithm to handle both multiple forward scattering and back scattering with light attenuation. This constitutes a complete system for shadow generation and light propagation.

Time-varying interval volumes

In this paper, we study the interval segmentation and direct rendering of time-varying volumetric data to provide a more effective and interactive volume rendering of time-varying structured and unstructured grids. Our segmentation is based upon intervals within the scalar field between time steps, producing a set of geometrically defined time-varying interval volumes. To construct the time-varying interval volumes, we cast the problem one dimension higher, using a five-dimensional isocontour construction for interactive computation or segmentation. The key point of this paper is how to render the time-varying interval volumes directly. We directly render the 4D interval volumes by projecting the 4D simplices onto 3D, decomposing the projected 4-simplices to 3-simplices and then rendering them using a modified hardware-implemented projected tetrahedron method. In this way, we can see how interval volumes change with the time in one view. The algorithm is independent of the topology of the polyhedral cells comprising the grid, and thus offers an excellent enhancement to the volume rendering of time-varying unstructured grids. Another advantage of this algorithm is that various volumetric and surface boundaries can be embedded into the time-varying interval volumes.

iSBVR: isosurface-AIDED hardware acceleration techniques for slice-based volume rendering

In this paper, we examine the performance of the early z-culling feature on current high-end commodity graphics cards and present an isosurface-aided hardware acceleration algorithm for slice-based volume rendering (iSBVR) to maximize its utilization. We analyze the computational models for early z-culling of the texture based volume rendering. We demonstrate that the performance improves with two to four times speedup against an original straightforward SBVR on an ATI 9800pro display board. As volumetric shaders become increasingly complex, the advantages of fast z-culling will become even more pronounced.

A new 3D display using a dynamically reconfigurable display matrix surface

This paper presents a new three-dimensional display system using a reconfigurable projection manifold. The display surface is warped interactively into a non-planar manifold according to the scene images depth information for the current view. Scene images are projected onto the display using computer graphics and a post-process warping. We currently use two projections to provide a complete coverage of the manifold surface from oblique projection angles. In this paper, we describe the overall design issues and goals of the projector-based rendering process, introduce the algorithms to model the projection manifold surface and describe the algorithms used to create the pre-warped images from the scene image. Our current prototype is a discrete tile-based approximation to the depth manifold. We describe our experiments with two such prototypes, one using a static test configuration, and another actuated using linear motion controllers for each discrete tile. This second prototype can be configured dynamically, as the depth information in the scene changes.

Fast dynamic flow volume rendering using textured splats on modern graphics hardware

Dynamic flow volume rendering of three-dimensional vector fields offers better insights into the continuum and dynamics of the data field under investigation. Consumer graphics cards have seen a rapid explosion of performance and capabilities over the past few years. This paper explores the development of the Textured Splats algorithm for direct flow volume rendering of vector fields, that utilizes this new hardware. This paper presents the technique using the new hardware features like vertex programs, OpenGL multi-textures and register combiner extensions to implement fast dynamic flow volume rendering on a PC equipped with an NVIDIA GeForce4 display card. Several anisotropic textured splats are investigated to implement flow volume rendering.

8 – Volume Rendering Using Splatting

The term “splat” means to spread flat or flatten out. It was first used for the deposition of a single 3D reconstruction kernel being projected and integrated onto the image screen. These were referred to as footprints, implying the extent to which a single voxel or reconstruction kernel covers the image plane. This chapter discusses some design goals that lead to the most effective kernels for splatting. In splatting, each voxel is represented by a 3D kernel weighted by the voxel value. The 3D kernels are integrated into a generic 2D footprint along the traversing ray from the eye. This footprint can be efficiently mapped onto the image plane; the final image is obtained by the collection of all projected footprints weighted by the voxel values. This splatting approach is fast, but it suffers from color bleeding and popping artifacts due to incorrect volume integration. The chapter also examines the illumination or volume rendering integral.

A tile-based 3D frame using a reconfigurable display matrix

Figure 1 shows the diagram of our 3D frame system. The scene image is divided into the same number of the tiles as the display matrix, pre-warped and projected from the left projector onto the non-planar display surface of the matrix. The displacement of each tile is determined by the average depth of the pixels that project onto it. Since the display surface is a discontinuous manifold, shadows occur due to the illumination from the left projector and may be visible to the user in the walking area. The right projector is used to illuminate or hide these shadow regions. The user at the center of the walking area sees a geometrically correct scene image on the non-planar display wall. Users, viewing the display off center, see a geometric approximation to the 3D model with per-pixel illumination. To change the display surface interactively, we have developed a robotic display matrix. Figure 2 shows our matrix with a 16×8 set of display tiles. Each tile is driven by an inexpensive servo motor. We can adjust the tiles’ displacement to create different scenes as well as provide interactivity or dynamic scene support by modifying the tiles’ displacement in real time.