Abstract: We present a first running video see-through augmented reality system on a consumer cell-phone. It supports the detection and differentiation of different markers, and correct integration of rendered 3D graphics into the live video stream via a weak perspective projection camera model and an OpenGL rendering pipeline.

Abstract: This paper presents a simple approach for rendering isosurfaces of a scalar field. Using the vertex programming capability of commodity graphics cards, we transfer the cost of computing an isosurface from the Central Processing Unit (CPU), running the main application, to the Graphics Processing Unit (GPU), rendering the images. We consider a tetrahedral decomposition of the domain and draw one quadrangle (quad) primitive per tetrahedron. A vertex program transforms the quad into the piece of isosurface within the tetrahedron (see Figure 2). In this way, the main application is only devoted to streaming the vertices of the tetrahedra from main memory to the graphics card. For adaptively refined rectilinear grids, the optimization of this streaming process leads to the definition of a new 3D space-filling curve, which generalizes the 2D Sierpinski curve used for efficient rendering of triangulated terrains. We maintain the simplicity of the scheme when constructing view-dependent adaptive refinements of the domain mesh. In particular, we guarantee the absence of T-junctions by satisfying local bounds in our nested error basis. The expensive stage of fixing cracks in the mesh is completely avoided. We discuss practical tradeoffs in the distribution of the workload between the application and the graphics hardware. With current GPU's it is convenient to perform certain computations on the main CPU. Beyond the performance considerations that will change with the new generations of GPU's this approach has the major advantage of avoiding completely the storage in memory of the isosurface vertices and triangles.

Abstract: Pre-integrated volume rendering is an effective technique for generating high-quality visualizations. The precomputed lookup tables used by this method are slow to compute and can not include truly pre-integrated lighting due to space constraints. The lighting for pre-integrated rendering is therefore subject to the same sampling artifacts as in standard volume rendering. We propose methods to speed up lookup table generation and minimize lighting artifacts. The incremental subrange integration method we describe allows interactive lookup table generation in O(n2) time without the need for approximation or hardware assistance. The interpolated preintegrated lighting algorithm eliminates discontinuities by linearly interpolating illumination along the view direction. Both methods are applicable to any pre-integrated rendering method, including cell projection, ray casting, and hardware-accelerated algorithms.

Abstract: Volume visualization using isosurface extraction is a well-researched topic. Research demonstrated that even for unstructured grids peak performances of millions of tetrahedra per second can be achieved by exploiting the parallel processing capabilities of modern GPUs. In this paper we present a hardware-accelerated solution that further improves the extraction performance. In contrary to existing approaches, our technique explicitly extracts the isosurface geometry in a fragment program by rendering only a single screen-sized quadrilateral. The extracted geometry is directly written to an on-board graphics memory object allowing for direct rendering without further bus transfers. Additionally, the geometry can be manipulated by shader programs or read back to the application for further processing. Examples and application scenarios are given that can benefit from our approach.

Abstract: Rendering GUI widgets with generic look and feel by receiving in a display device a master definition of a graphics display, the master definition including at least one graphics definition element, the graphics definition element including a reference to a protowidget and one or more instance parameter values characterizing an instance of the protowidget, the protowidget includes a definition of a generic GUI object, including generic display values affecting overall look and feel of the graphics display, and rendering at least one instance of the protowidget to a graphics display in dependence upon the generic display values and the instance parameter values.

Abstract: Free viewpoint video provides the possibility to freely navigate within dynamic real world video scenes by choosing arbitrary viewpoints and view directions. So far, related work only considered free viewpoint video extraction, representation, and rendering methods. Compression and transmission has not yet been studied in detail and combined with the other components into one complete system. In this paper, we present such a complete system for efficient free viewpoint video extraction, representation, coding, and interactive rendering. Data representation is based on 3D mesh models and view-dependent texture mapping using video textures. The geometry extraction is based on a shape-from-silhouette algorithm. The resulting voxel models are converted into 3D meshes that are coded using MPEG-4 SNHC tools. The corresponding video textures are coded using an H.264/AVC codec. Our algorithms for view-dependent texture mapping have been adopted as an extension of MPEG-4 AFX. The presented results illustrate that based on the proposed methods a complete transmission system for efficient free viewpoint video can be built.

Abstract: Artistic rendering is an important research area in Computer Graphics, yet relatively little attention has been paid to the projective properties of computer generated scenes. Motivated by the surreal storyboard of an animation in production---Ryan---this paper describes interactive techniques to control and render scenes using nonlinear projections. The paper makes three contributions. First, we present a novel approach that distorts scene geometry such that when viewed through a standard linear perspective camera, the scene appears nonlinearly projected. Second, we describe a framework for the interactive authoring of nonlinear projections defined as a combination of scene constraints and a number of linear perspective cameras. Finally, we address the impact of nonlinear projection on rendering and explore various illumination effects. These techniques, implemented in Maya and used in the production of the animation Ryan, demonstrate how geometric and rendering effects resulting from nonlinear projections can be seamlessly introduced into current production pipelines.

Abstract: With the tremendous advances in both hardware capabilities and rendering algorithms, rendering performance is steadily increasing. Even consumer graphics hardware can render many million triangles per second. However, scene complexity seems to be rising even faster than rendering performance, with no end to even more complex models in sight. In this paper, we are targeting the interactive visualization of the "Boeing 777" model, a highly complex model of 350 million individual triangles, which - due to its sheer size and complex internal structure – simply cannot be handled satisfactorily by today's techniques. To render this model, we use a combination of real-time ray tracing, a low-level out of core caching and demand loading strategy, and a hierarchical, hybrid volumetric/lightfield-like approximation scheme for representing not-yet-loaded geometry. With this approach, we are able to render the full 777 model at several frames per second even on a single commodity desktop PC.

Abstract: A system and method for optimizing the performance of a graphics intensive software program for graphics acceleration hardware. This system and method encompasses a procedure that validates the different functions of a 3D acceleration capable video card, decides whether to use the acceleration hardware and optimizes the software application to selectively use the functions that work on the specific video acceleration card. Functions checked include sub-pixel positioning, opacity, color replacement and fog. If these tests are successful, then the graphics acceleration is used by the software application. However, if the tests are not successful the decision is made not to use graphics accelerator. Those with ordinary skill in the art will realize that it is not necessary to perform all of the tests in a specific order. Additionally, other types of tests could be performed to ensure software application and video card compatibility before the software application is uses graphics acceleration to render 3D graphics.

Abstract: Coherence of displayed images is provided for a graphics processing systems having multiple processors operating to render different portions of a current image in parallel. As each processor completes rendering of its portion of the current image, it generates a local ready event, then pauses its rendering operations. A synchronizing agent detects the local ready event and generates a global ready event after all of the graphics processors have generated local ready events. The global ready signal is transmitted to each graphics processor, which responds by resuming its rendering activity.

Abstract: The Fourier volume rendering technique operates in the frequency domain and creates line integral projections of a 3D scalar field. These projections can be efficiently generated in ) log O( 2 N N time by utilizing the Fourier Slice-Projection theorem. However, until now, the mathematical difficulty of the Fast Fourier Transform prevented acceleration by graphics hardware and therefore limited a wider use of this visualization technique in state-of-theart applications. In this paper we describe how to utilize current commodity graphics hardware to perform Fourier volume rendering directly on the GPU. We present a novel implementation of the Fast Fourier Transform: This Split-Stream-FFT maps the recursive structure of the FFT to the GPU in an efficient way. Additionally, highquality resampling within the frequency domain is discussed. Our implementation visualizes large volumetric data set in interactive frame rates on a mid-range computer system.

Abstract: This paper describes a perceptually-based image comparison process that can be used to tell when images are perceptually identical even though they contain some numerical differences. The technique has shown much utility in the production testing of rendering software.

Abstract: A platform-independent occlusion culling library for dynamic environments, dPVS, can benefit such applications as CAD and modeling tools, time-varying simulations, and computer games. Visibility optimization is currently the most effective technique for improving rendering performance in complex 3D environments. The primary reason for this is that during each frame the pixel processing subsystem needs to determine the visibility of each pixel individually. Currently, rendering performance in larger scenes is input sensitive, and most of the processing time is wasted on rendering geometry not visible in the final image. Here we concentrate on real-time visualization using mainstream graphics hardware that has a z-buffer as a de facto standard for hidden surface removal. In an ideal system only the complexity of the geometry actually visible on the screen would significantly impact rendering time - 3D application performance should be output sensitive.

Abstract: A method and apparatus for rendering three-dimensional graphics using a streaming render-cache with a multi-threading, multi-core graphics processor are disclosed. The graphics processor includes a streaming render-cache and render-cache controller to maintain the order in which threads are dispatched to the graphics engine, and to maintain data coherency between the render-cache and the main memory. The render-cache controller blocks threads from being dispatched to the graphics engine out of order by only allowing one sub-span to be in-flight at any given time.

Abstract: While interactive visualization of rectilinear gridded volume data sets can now be accomplished using texture mapping hardware on commodity PCs, interactive rendering and exploration of large scattered or unstructured data sets is still a challenging problem. We have developed a new approach that allows the interactive rendering and navigation of procedurally-encoded 3D scalar fields by reconstructing these fields on PC class graphics processing units. Since the radial basis functions (RBFs) we use for encoding can provide a compact representation of volumetric scalar fields, the large grid/mesh traditionally needed for rendering is no longer required and ceases to be a data transfer and computational bottleneck during rendering. Our new approach will interactively render RBF encoded data obtained from arbitrary volume data sets, including both structured volume models and unstructured scattered volume models. This procedural reconstruction of large data sets is flexible, extensible, and can take advantage of the Moore's Law cubed increase in performance of graphics hardware.

Abstract: This paper presents an interactive non-photorealistic rendering (NPR) system that stylizes and renders outdoor scenes captured by 3D laser scanning. In order to deal with the large size, complexity and inherent incompleteness of data obtained from outdoor scans, our system represents outdoor scenes using points instead of traditional polygons. Algorithms are then developed to extract, stylize and render features from this point representation. In addition to conveying various NPR styles, our system also promises consistency in animation by maintaining stroke coherence and density. We achieve NPR of large data at interactive rates by designing novel data structures and algorithms as well as leveraging new features of commodity graphics hardware.

Abstract: Pattern recognition and computer vision tasks are computationally intensive, repetitive, and often exceed the capabilities of the CPU, leaving little time for higher level tasks. We present a novel computer architecture which uses multiple commodity computer graphics devices to perform pattern recognition and computer vision tasks many times faster than the CPU. This is a parallel computing architecture that is quickly and easily constructed from the readily available hardware. It is based on parallel processing done on multiple graphics processing units (GPUs). An eigenspace image recognition approach is implemented on this parallel graphics architecture. This paper discusses methods of mapping computer vision algorithms to run efficiently on multiple graphics devices to maximally utilize the underlying graphics hardware. The additional memory and memory bandwidth provided by the graphics hardware provided for significant speedup of the eigenspace approach. We show that graphics devices parallelize well and provide significant speedup over a CPU implementation, providing an immediately constructible low cost architecture well suited for pattern recognition and computer vision.

Abstract: We present a simple point-based multiresolution structure for interactive visualization of very large point sampled models on consumer graphics platforms. The structure is based on a hierarchy of precomputed object-space point clouds. At rendering time, the clouds are combined coarse-to-fine with a top-down structure traversal to locally adapt sample densities according to the projected size in the image. Since each cloud is made of a few thousands of samples, the multiresolution extraction cost is amortized over many graphics primitives, and host-to-graphics communication effectively exploits on-board caching and object based rendering APIs. The progressive block based refinement nature of the rendering traversal is well suited to hiding out-of-core data access latency, and lends itself well to incorporate backface, view frustum, and occlusion culling, as well as compression and viewdependent progressive transmission. The resulting system allows rendering of complex models at high frame rates (over 60M splat/second), supports network streaming, and is fundamentally simple to implement.

Abstract: We describe an approach for rendering large terrains in real-time. A digital elevation map defines the rough shape of the terrain. During rendering, procedural geometric and texture detail is added by the graphics hardware. We show, how quad meshes can be generated quickly that have a locally varying resolution that is optimized for the inclusion of procedural detail. We obtain these distorted meshes by importance based warping of geometry images. The resulting quad mesh can then be rendered very efficiently by graphics hardware, which also adds all visible procedural detail using vertex and fragment programs.

Abstract: A server computer hosts one or more application programs that are accessed by a client computer. Higher-level graphics commands describing graphics images are received from the application programs. The server computer determines whether the client computer is able to generate graphics using the higher-level graphics commands or generates graphics using relatively lower-level graphics commands. The server computer sends higher-level or relatively lower-level graphics commands depending on whether the client computer generates graphics using higher-level or relatively lower-level graphics commands.

Abstract: A system and method for optimizing power usage and performance during data processing. A multi-processor graphics processing system includes a low power graphics processor and a high performance graphics processor. When a low power condition exists only the low power graphics processor is used to process graphics data and the high performance graphics processor is turned off. When turned off, the high performance graphics processor does not consume either static or dynamic power. When the low power condition does not exist, the high performance graphics processor is turned on and the low power graphics processor and the high performance graphics processor are used to process the graphics data.

Abstract: We present a technique that allows the implementation of a stack on programmable graphics hardware, using textures and fragment shaders. This development enables a whole new class of GPU algorithms, including recursive functions on complex data structures. Kd-tree traversal for ray tracing is demonstrated as an application. The traversal core was integrated into a purely GPU based photorealistic path tracing system.

Abstract: The only surface primitives that are supported by common graphics hardware are triangles and more complex shapes have to be triangulated before being sent to the rasterizer. Even quadrilaterals, which are frequently used in many applications, are rendered as a pair of triangles after splitting them along either diagonal. This creates an undesirable C1 -discontinuity that is visible in the shading or texture signal. We propose a new method that overcomes this drawback and is designed to be implemented in hardware as a new rasterizer. It processes a potentially non-planar quadrilateral directly without any splitting and interpolates attributes smoothly inside the quadrilateral. This interpolation is based on a recent generalization of barycentric coordinates that we adapted to handle perspective correction and situations in which a quadrilateral is partially behind the point of view.

Abstract: This paper presents a novel framework for illustrating surfaces in a volume. Surfaces are illustrated by drawing only feature lines, such as silhouettes, valleys, ridges, and surface hatching strokes, and are embedded in volume renderings. This framework promises effective illustration of both surfaces and volumes without occluding or cluttering each other. A two-step approach has been taken: the first step depicts surfaces; the second step performs volume rendering, at the same time embedding surfaces from the first step.We introduce Procedurally Perturbed Image Processing (PIP), a new method for enhancing both feature detection and depiction of surfaces. We also present implementation strategies, especially those leveraging modern graphics hardware, for delivering an interactive rendering system. Our implementation results have shown that this mixed form of rendering improves volume visualization and is efficient.

Abstract: The increasing popularity of points as rendering primitives has led to a variety of different rendering algorithms, and in particular the different implementations compare like apples to oranges In this paper we revisit a number of recently developed point-based rendering implementations We briefly summarize a few proposed hierarchical multiresolution point data structures and their advantages Based on a common multiresolution framework we then describe and examine different hardware accelerated point rendering algorithms Experimental results are given with respect to performance timing and rendering quality for the different approaches

Abstract: We present a real-time algorithm for rendering volumetric 3D Magic Lenses™ having arbitrary convex shapes. During fragment processing the algorithm performs a second depth test using a shadow map. Exploiting the second depth test we are able to classify each fragment, with respect to its position relative to the lens volume. Using this classification we first render the geometry behind the lens volume, then the geometry intersecting the lens volume using a different visual appearance and finally the parts in front of the lens volume. Regardless of the shape of the lens volume just two additional rendering passes are needed. Furthermore there are no limitations to the choice of visual appearance used to enhance expressiveness of the virtual world. We will describe theoretical and practical aspects of the algorithm and our implementation, which is accelerated by current graphics hardware.

Abstract: In this paper, we present an empirical approach for rendering realistic watercolor effects in real-time While watercolor being a versatile media, several characteristics of its effects have been categorized in the past We describe an approach to recreate these effects using the Kubelka-Munk compositing model and the Sobel filter Using modern per-pixel shading hardware, we present a method to render these effects in an interactive frame-rate