25 results
Search Results
Now showing 1 - 10 of 25
Item Light Field Rendering for Games(The Eurographics Association, 2008) Todt, S.; Rezk-Salama, C.; Kolb, A.; Ik Soo Lim and Wen TangThis paper studies the problem of integrating high-quality light field rendering into state-of-the-art real-time computer games. We present a complete production pipeline for generation of light fields from arbitrary complex 3D content using commercial 3D modeling software commonly used in the gaming industry. We show how recent advances in light field rendering techniques can be used to composite light fields with dynamic scene content. The results demonstrate the potential of light fields as a valuable extension to polygonal rendering techniques by employing accurate silhouette reconstruction and correct per-pixel depth values for scene composition. Level of detail techniques are applied to light fields to optimize rendering performance. Dynamic light field rendering is implemented to account for animation, deformation and varying lighting conditions.Item Generation and Tessellation of Tree Stems(The Eurographics Association, 2008) Skjermo, Jo; Ik Soo Lim and Wen TangWhen visualizing tree stems and branches for use in interactive applications, the polygon models resolution are usually as low as possible to achieve a high frame rate. Also, to ease animation and mesh generation, each branch of a tree model is often considered as a distinct mesh. However, by using a single watertight mesh for a tree, together with (GPU-based) tessellation, both the resolution and appearance of a tree can be greatly improved while maintaining a high frame rate. This paper presents concepts, ideas and early work on generating watertight polygon meshes of animated trees stems suitable for refinement and tessellation of such meshes.Item Contouring with Uncertainty(The Eurographics Association, 2008) Osorio, R. S. Allendes; Brodlie, K. W.; Ik Soo Lim and Wen TangThe visualization of uncertainty remains one of the major challenges for the visualization community. To achieve this, we need to understand and develop methods that allow us not only to consider uncertainty as an extra variable within the visualization process, but to treat it as an integral part. In this paper, we take contouring, one of the most widely used visualization techniques for two dimensional data, and focus on extending the concept of contouring to uncertainty. We develop special techniques for the visualization of uncertain contours. We illustrate the work through application to a case study in oceanography.Item Localised Topology Correction for Hypertextured Terrains(The Eurographics Association, 2008) Gamito, Manuel N.; Maddock, Steve C.; Ik Soo Lim and Wen TangTerrains for computer graphics have traditionally been modelled with height fields. In the case of procedurally defined terrains, the height field is generated by a displacement map zItem Towards Image Rendering using models of Image Manifolds(The Eurographics Association, 2008) Woodland, A.; Labrosse, F.; Ik Soo Lim and Wen TangBuilding computer representations of real objects, which could be archaeological artifacts or planned buildings, is a very active research area. Typically objects are represented in computer memory as a collection of geometric primitives. In this paper we build on the concept of Image Manifolds, which we model in image space. This enables us to model the appearance of an object essentially as a collection of images. We discuss one possible approach to modelling image manifolds, which uses partial differential equations.Item Self-Organising Maps for Implicit Surface Reconstruction(The Eurographics Association, 2008) Yoon, Mincheol; Ivrissimtzis, Ioannis; Lee, Seungyong; Ik Soo Lim and Wen TangThis paper proposes an implicit surface reconstruction algorithm based on Self-Organising Maps (SOMs). The SOM has the connectivity of a regular 3D grid, each node storing its signed distance from the surface. At each iteration of the basic algorithm, a new training set is created by sampling regularly along the normals of the input points. The main training iteration consists of a competitive learning step, followed by several iterations of Laplacian smoothing. After each training iteration, we use extra sample validation to test for overfitting. At the end of the training process, a triangle mesh is extracted as the zero level set of the SOM grid. Validation tests and experiments show that the algorithm can cope with the noise of raw scan data. Timing measurements and comparisons show that the algorithm is fast, because the fixed and regular connectivity of the SOM means that the search of the node nearest to a sample can be done efficiently.Item Interactive Grid Based Binning for Information Visualization(The Eurographics Association, 2008) Longshaw, S. M.; Turner, M. J.; Hewitt, W. T.; Ik Soo Lim and Wen TangClutter within information visualization (infovis) systems is an area of continuing concern and is receiving increasing research interest. Solutions to the problem vary in their approach, ranging from novel visualizations designed specifically to cope with high data density, through to statistical methodologies such as binning. This paper presents a flexible method that allows interactive placement of a grid based binning system that aims to enhance traditional information visualization techniques. User evaluations employing two specific visualization methods are described using a prototype grid based binning system. The method is shown to be a quick and easy way to visually segment a data domain, while the two visualization techniques presented are shown to provide effective data overview. Due to the abstracted nature of the binning grid, its applicability goes beyond the examples provided in this paper, therefore it could be considered as a generic data reduction and/or overview technique within many systems.Item Time-varying Image Data Visualization Framework for Application in Cardiac Catheterization Procedures(The Eurographics Association, 2008) Ma, Ying Liang; Rhode, Kawal S.; King, Andy P.; Gao, Gang; Chinchapatnam, Phani; Schaeffter, Tobias; Razavi, Reza; Saetzler, Kurt; Ik Soo Lim and Wen TangVisualization plays an important role in image guided surgery. This paper presents a real-time 3D motion visualization method where pre-computed meshes of the beating heart are synchronized with and overlaid onto live X-ray images. This provides the surgeon with a navigational aid in guiding catheters during cardiac catheterization. In order to generate time-varying meshes of the beating heart, we first acquire a time-series of images of the patient using Magnetic Resonance Imaging (MRI). The MRI heart images used for the cardiac catheterization procedures can either be contrast-enhanced by injecting a contrast agent prior to imaging or they can be unenhanced. The contrast-enhanced images can easily be segmented and binarized using a fixed grey-level threshold. In this case, we can use an adaptive Delaunay-based surface extraction algorithm for mesh generation, for which specifically developed for noisy binary image data sets. For unenhanced images, we have to choose a semi-automated segmentation approach, where a region of interest in the patient's heart is outlined manually in an intermediate slice in the 3-D MRI data set and then propagated to neighbouring slices. In a next step, the extracted snake contours are propagated in time from the first phase of the cardiac cycle to subsequent phases using multiple snake contours. In this scenario, the final mesh is generated using a serial section reconstruction algorithm. However, due to the nature of the underlyling MRI images which frequently contain areas of inhomogenous contrast caused by motion and blood flow, it is difficult to generate a smooth mesh directly from the result of the previously described semi-automatic segmentation procedure. Therefore, we also introduce a contour-based mesh smoothing algorithm using a 1D Gaussian filter in order to post-process the snake contours along the series of cross-sections before reconstruction.Item A Benchmarking Framework for Static Collision Detection(The Eurographics Association, 2008) Diktas, Engin Deniz; Sahiner, Ali Vahit; Ik Soo Lim and Wen TangPerformanceof static collision detection queries depends on the type of the hierarchy chosen as well as the relative positioning of the colliding objects. In order to evaluate the performance of bounding volume hierarchies, relevant criteria that affect the query performance need to be determined and the sample space should be generated accordingly. In this paper we present a benchmarking framework for evaluating the performance of various static collision detection algorithms. In this framework, instances of a moving rigid object are placed on the surface of another instance of the same object fixed at a certain position, where the contact occurs for the first time. Then by offsetting the surface inwards (outwards) we generate new surfaces that are at a certain fixed negative (positive) distance to the original surface. Placing the moving object on these offset surfaces makes the object penetrate (approach) the fixed object at a fixed distance. For offset surface generation we create a signed distance field and run marching cubes algorithm on it.Item A Physical Model for the Polarized Scattering of Light(The Eurographics Association, 2008) Brayford, David; Turner, Martin; Hewitt, W. T.; Ik Soo Lim and Wen TangThe change in polarization state due to the interaction of light with the surface and beneath the surface of an object has become increasingly important in realistic image synthesis of materials such as metallic, iridescent and pearlescent paint, skin, hair and cosmetics. This paper presents a model for the anisotropic scattering of polarized light based upon the physics of light; which is capable of calculating both partial and complete polarization using a combination of Jones and Mueller calculus, as well as incorporating self shadowing effects.
- «
- 1 (current)
- 2
- 3
- »