WSCG '2008: Communication Papers
Permanent URI for this collection
Browse
Recent Submissions
Item An Analysis of Cache Awareness for Interactive Selective Rendering(Václav Skala - UNION Agency, 2008) Dubla, Piotr; Chalmers, Alan; Debattista, Kurt; Cunningham, Steve; Skala, VáclavInteractive high-fidelity rendering is one of the major goals of computer graphics. Algorithms based on ray tracing are usually used to drive high-fidelity renderers. While ray tracing is often thought to be impractical for real time performance, recent algorithmic advances in the field have led to the development of interactive ray tracers which leverage the performance of modern parallel systems and cache awareness to obtain real-time rates for moderately complex scenes using Whitted-style ray tracing. Another method used for accelerating ray tracing is the use of selective rendering where only the pixels that need to be computed are traced and the remainder are computed by other means such as interpolation. The choice of which pixels to render may depend on a number of factors, from simple ones that just compare the radiance values of the rays within a certain area and shoot further rays recursively if the difference in radiance is above a certain threshold, to more complex ones based on visual attention. Selective rendering algorithms may not be perfectly compatible with current interactive ray tracing techniques since selective rendering methods tend to be naturally incoherent, and the computation of rays at a stride may result in expensive cache misses. In this paper we analyse the effect of selective rendering on interactive ray tracing and hint at possible solutions that would allow selective rendering to be compatible with interactive rendering methods to further improve rendering speeds.
Item Controlling GPU-based Volume Rendering using Ray Textures(Václav Skala - UNION Agency, 2008) Raspe, Matthias; Müller, Steffan; Cunningham, Steve; Skala, VáclavIn this paper we introduce a novel approach to control different rendering parameters of volume ray casting. Since the introduction of ray casting implementations on programmable graphics hardware, both performance and flexibility have increased and are able to outperform texture-based techniques. In addition, by using rays for computing the volume integral instead of proxy geometry one has more control over local settings. Therefore, we employ dependent texture lookups to user editable 2D textures, thus allowing for interactive parameter setting on a per-ray basis, at a negligible performance overhead on modern graphics hardware. By those means we are able to control the volume rendering in a way not possible with proxy-based direct volume rendering and demonstrate some exemplary uses.Item 5 Axis Milling Simulation using a Swept Volume via Gauss Map(Václav Skala - UNION Agency, 2008) Lee, Seok Won; Nestler, Andreas; Cunningham, Steve; Skala, VáclavIn this paper, a novel approach to 5 axis NC milling simulation using GPU is presented. Modelling swept volumes (SV) of a milling tool that undergoes a translational and rotational motion is based on the idea behind Gauss maps. The main process is that it 1) transforms the tool shape and the moving direction of the tool from Euclidean space into an unit sphere by means of Gauss function and gets the spherical domains which are called Tool map (T-Map) and Contact map (C-Map) respectively 2) finds an intersection region of T-Map with C-Map and 3) transforms intersected region inversely into Euclidean space to get the envelope profile of the tool at prescribed instant time. As of now, Graphics Processing Unit (GPU) in graphics hardware is being developed more drastically as compared with Central Processing Unit (CPU). To achieve our purpose of updating the work piece GPU is exploited as follows: 1) SV is rendered using graphics hardware (Model building) 2) color and depth information is read back from the frame buffer in GPU to CPU (sampling) 3) at last, the updated work piece is built based on extracted data (surface reconstruction).Item Accelerating Rendering of NURBS Surfaces by Using Hybrid Ray Tracing(Václav Skala - UNION Agency, 2008) Abert, Oliver; Bröcker, Markus; Spring, Rafael; Cunningham, Steve; Skala, VáclavIn this paper we present a new method for accelerating ray tracing of scenes containing NURBS (Non Uniform Rational B-Spline) surfaces by exploiting the GPU’s fast z-buffer rasterization for regular triangle meshes. In combination with a lightweight, memory efficient data organization this allows for fast calculation of primary ray intersections using a Newton Iteration based approach executed on the CPU. Since all employed shaders are kept simple the algorithm can profit from older graphics hardware as well. We investigate two different approaches, one initiating ray-surface intersections by referencing the surface through its child-triangles. The second approach references the surface directly and additionally delivers initial guesses, required for the Newton Iteration, using graphics hardware vector interpolation capabilities. Our approaches achieve a rendering acceleration of up to 95% for primary rays compared to full CPU ray tracing without compromising image quality.Item VDesktop: Event Management and Physically Based Behaviour in Tabletop Displays(Václav Skala - UNION Agency, 2008) Moreno, Aitor; Heukamp, Jan; Posada, Jorge; García-Alonso, Alejandro; Cunningham, Steve; Skala, VáclavApplications for digital tabletops have some notable differences when compared to traditional desktop applications, the principal difference being that the user input method is not the traditional combination of mouse and keyboard. This work addresses the difficulties and characteristics of event-handling management, when the applications are extended from 2D to 3D in the context of tabletop displays and where the interactions between the user and the represented 3D objects are more complex. This complexity increases when physically based behaviour of the objects is considered. More specifically, the scope of this work is oriented to the support of physic-based simulation events in a tabletop display with an implementation of a system based on the OGRE graphical library and the ODE physical library, that are able to handle such elements.Item Interactive Images using Illustration Watermarks: Techniques, Studies, and Applications(Václav Skala - UNION Agency, 2008) Sonnet, Henry; Unger, Andrea; Schumann, Heidrun; Cunningham, Steve; Skala, VáclavStatic 2D images can be turned into interactive media by providing descriptive information that is shown on demand. For this purpose, two challenges need to be solved. First, appropriate techniques for the storage of descriptive information and its linkage to the image content are needed. We address this challenge by applying techniques from information hiding, a procedure we call Illustration Watermarking. In a user study, a new adaptive approach optimized for our application outperformed a traditional proceeding in terms of quality. The second challenge is to design interaction procedures which support the user in the exploration of descriptive information. Therefore, concepts are proposed to let the user find and retrieve the embedded data and its properties. Finally, we present various application scenarios.Item Accelerated Stereoscopic Rendering using GPU(Václav Skala - UNION Agency, 2008) de Sorbier, François; Nozick, Vincent; Biri, Venceslas; Cunningham, Steve; Skala, VáclavThis paper presents a new method to create a pair of stereoscopic images in one pass. Our algorithm takes advantage of the latest version of GPUs, including geometry shaders. Given the left viewpoint of a scene, primitives are duplicated and transformed to compute the right image. Hence this method saves the extra-time required to recompute attributes of the vertices for the second view in the traditional rendering pipeline. Even for very complex scenes, our method provides stereoscopic pair roughly twice faster than a traditional method and involves few additional implementations.Item Keypatches: A New Type of Local Features for Image Matching and Retrieval(Václav Skala - UNION Agency, 2008) Śluzek, Andrzej; Cunningham, Steve; Skala, VáclavThe paper presents a new approach in defining (and detecting, as an additional option) local features that can be used for matching images and/or visual information retrieval. The method is based on the moment-based pattern detectors presented in previous papers. The proposed local features (preliminarily called keypatches) are obtained by approximating circular windows located at keypoints pre-detected using any typical detector with a selection of geometric patterns. At each keypoint, the optimum approximations (for all available patterns) of the window are computed using moment-based equations. For any approximation, its similarity to the actual window content is estimated. The keypatch is defined if a sufficiently accurate approximation exists. Keypoints where the window cannot be approximated with the sufficient accuracy are ignored. If no pre-detector of keypoints is available, the method itself can find the initial locations of the keypoints. The proposed approach is suitable for both grey-level and colour images (though the latter are only briefly discussed in the paper). Exemplary results explaining the method and illustrating its performances are included and discussed.Item An Image-Based Approach to Visual Feature Space Analysis(Václav Skala - UNION Agency, 2008) Schreck, Tobias; Schneidewind, Jörn; Keim, Daniel A.; Cunningham, Steve; Skala, VáclavMethods for management and analysis of non-standard data often rely on the so-called feature vector approach. The technique describes complex data instances by vectors of characteristic numeric values which allow to index the data and to calculate similarity scores between the data elements. Thereby, feature vectors often are a key ingredient to intelligent data analysis algorithms including instances of clustering, classification, and similarity search algorithms. However, identification of appropriate feature vectors for a given database of a given data type is a challenging task. Determining good feature vector extractors usually involves benchmarks relying on supervised information, which makes it an expensive and data dependent process. In this paper, we address the feature selection problem by a novel approach based on analysis of certain feature space images. We develop two image-based analysis techniques for the automatic discrimination power analysis of feature spaces. We evaluate the techniques on a comprehensive feature selection benchmark, demonstrating the effectiveness of our analysis and its potential toward automatically addressing the feature selection problem.Item Human Computer Interactions and Task Performance(Václav Skala - UNION Agency, 2008) Nassiri, Nasser; Powell, Norman; Moore, David; Cunningham, Steve; Skala, VáclavAs humans start to spend more time in collaborative virtual environment (CVE), coordinating the interaction between the humans in these environments is becoming increasingly important. We have been investigating one aspect of such coordination, namely the issue of an avatar’s “personal space”. Intuitively it can be expected that CVE users might decrease their task performance when their avatar personal space is invaded since this socially unacceptable act tends to cause anxiety. To investigate the effect of personal space invasion on a user’s task performance, we have conducted a controlled experiment measuring the effect of personal space invasion on a user’s task performance. The results of the experiment suggest that a user whose personal space is invaded performs more slowly than a user whose avatar’s personal space is not invaded.Item 3D Marker Data For Motion Recognition(Václav Skala - UNION Agency, 2008) Köhler, Hildegard; Pruzinec, Martin; Feldmann, Tobias; Wörner, Annika; Cunningham, Steve; Skala, VáclavAccurate human motion models are a prerequisite for most applications dealing with the tracking, reconstruction or recognition of human motions. Often a uniform model is used, approximating the average of the evaluated subjects. However we expect most applications can be improved by using individual models for each subject with its personal body masses and features. Hence we propose an algorithm for automatic reconstruction of anatomical features of subjects from labeled 3D marker data by a parameterized generic model. Our main contribution is a novel approach for automatically estimating skeletons of individual subjects and to transform them to a human body model by preserving its relative configuration. We show that a more accurate model can help in context of motion recognition by improving standard motion reconstruction with regard to its quantity and quality.Item Camera Calibration Methods Evaluation Procedure for Images Rectification and 3D Reconstruction(Václav Skala - UNION Agency, 2008) Guerchouche, Rachid; Coldefy, François; Cunningham, Steve; Skala, VáclavCamera calibration and images rectification are two necessary steps in most 3D reconstruction methods using image acquisition. This paper proposes an evaluation procedure for camera calibration methods for the case of 3D reconstruction using rectified multi-stereo images. The evaluation is based on the accuracy of the rectification and of the 3D reconstruction which are directly related to the calibration precision. Three methods are thus compared: Faugeras-Toscani, Zhang and a robust calibration algorithm. The procedure can be applied for computer vision systems with an arbitrary number of cameras and for any other calibration method. We show that, although the three methods provide significantly different intrinsic and stereo system parameter estimations, the rectified images of the planar target that we use for evaluation are relatively coherent and lead to close 3D reconstruction errors.Item Toolkit for registration and evaluation for 3D laser scanner acquisition(Václav Skala - UNION Agency, 2008) Synave, Rémi; Desbarats, Bats; Gueorguieva, Stefka; Cunningham, Steve; Skala, VáclavThe construction of numerical representation of 3D objects based on laser scan acquisition is composed by three steps: the range image acquisition, the registration and the integration. Experiments show that registration plays a key role in the final representation precision. This article is a practical report containing the description of a toolkit for registration and for evaluation of the registered mesh model according to geometrical and topological criteria. Examples of manipulation of the toolkit are provided for synthetic and real objects.Item Radon and Mojette Projections’ Equivalence for Tomographic Reconstruction using Linear Systems(Václav Skala - UNION Agency, 2008) Recur, Benoît; Desbarats, Pascal; Domenger, Jean-Philippe; Cunningham, Steve; Skala, VáclavIn medical imaging, the Radon transform, used for tomographic reconstruction, recovers a N-Dimensional image from a set of acquired (N-1)D projections. However, it implies approximations in discrete domain. Beside this transform, Mojette transform has been developped as a discrete and exact transform. However, its construction is incompatible with physical X-Rays properties. In this paper, after having recalled generalities about both transforms, we introduce the conditions under which their projections are equivalent and how to reconstruct an image with the Mojette backprojection from a Radon acquisition using linear systems. This method will be applied on some examples and results will be compared with usual Radon transform.Item A New Computer-aided Technique for Planning the Aesthetic Outcome of Plastic Surgery(Václav Skala - UNION Agency, 2008) Bottino, Andreas; Laurentini, Aldo; Rosano, Luisa; Cunningham, Steve; Skala, VáclavPlastic surgery plays a major role in today health care. Planning plastic face surgery requires dealing with the elusive concept of attractiveness for evaluating feasible beautification of a particular face. The existing computer tools essentially allow to manually warp 2D images or 3D face scans, in order to produce images simulating possible surgery outcomes. How to manipulate faces, as well as the evaluation of the results, are left to the surgeon’s judgement. We propose a new quantitative approach able to automatically suggest effective patient-specific improvements of facial attractiveness. The general idea is to compare the face of the patient with a large database of attractive faces, excluding the facial feature to be improved. Then, the feature of the faces more similar is applied, with a suitable morphing, to the face of the patient. In this paper we present a first application of the general idea in the field of nose surgery. Aesthetically effective rhinoplasty is suggested on the base of the entire face profile, a very important 2D feature for rating face attractiveness.Item Hybrid Terrain Shadow Ray Casting(Václav Skala - UNION Agency, 2008) Sakalauskas, Tomas; Cunningham, Steve; Skala, VáclavThis paper describes an efficient algorithm for real-time terrain shadowing by directional light. The main concept for this algorithm is fast preprocessing stage in model space that dramatically speeds up shadow calculation for pixels in rasterization phase. Preprocessing is very fast and can be performed every frame still achieving interactive frame rates. Therefore it is suitable for dynamic height field and moving light source visualization.Item Parallel 3D Split and Merge Segmentation with Oriented Boundary Graph(Václav Skala - UNION Agency, 2008) Baldacci, F.; Desbarats, P.; Cunningham, Steve; Skala, VáclavIn this paper, we present how 3D split and merge segmentation using topological and geometrical structuring with an Oriented Boundary Graph may be optimized by parallel algorithms. This structuring allows to implement efficiently split and merge operations, but since these treatments have often to be applied with large images, we have studied how to improve performances by parallelizing this process. After a short description of the structuring model and its construction, we describe algorithms for parallelizing the construction of the structuring and describe how this model can be maintained while using parallel processes. We explain the way of partitioning data for use with multiprocessor systems, and extension for use with NUMA architectures and graphics processing units (GPU) is described. Exemples on two medical images of different sizes is presented and execution time will be given.Item Dynamic Gestural Interaction with Immersive Environments(Václav Skala - UNION Agency, 2008) Malerczyk, Cornelius; Cunningham, Steve; Skala, VáclavThis paper describes our ongoing research work on deviceless interaction using hand gesture recognition with a calibrated stereo system. Video-based interaction is one of the most intuitive kinds of Human-Computer- Interaction with Virtual-Reality applications due to the fact that users are not wired to a computer. This kind of interaction should therefore be considered as the preferred kind of interaction especially for technically unversed users. Especially, if interaction with three-dimensional environments is considered, pointing is one of the most intuitive kind of interaction used by humans. Nevertheless, using a pointing posture solely lacks of some classical interaction metaphors like moving objects in 3D space. As the most important additional gestures beside the recognition of a pointing gesture, 'grab' and 'release' gestures has been identified to enable an interactive movement of 3D objects in virtual worlds. This paper describes our video-based gesture recognition system that uses two calibrated cameras observing the user in front of a large displaying screen, identifying three different hand gestures in real time and determining 3D information like the 3D position of the user's hand or the pointing direction if performed. Different scenario applications like a virtual chess game against the computer or an industrial scenario (placing filters to an air pump system in 3D space) were developed and tested.Item Estimation of Projector Defocus Blur by Extracting Texture Rich Region in Projection Image(Václav Skala - UNION Agency, 2008) Oyamada, Yuji; Saito, Hideo; Cunningham, Steve; Skala, VáclavTo use a projector anytime anywhere, a lot of projector-camera based approaches have been proposed. Focal correction technique, one of them, reduces the effect of defocus blur which occurs when a screen is located out of projector's depth-of-field. To achieve focal correction, we have to estimate the degree of defocus blur on the displayed images. Some previous methods use any special pattern images. A method which needs no special image has been proposed, however, this method can't estimate at less-textured or already blurred region in the projection image. In this paper, we propose a novel method for estimating PSF (Point Spread Function) without projecting any special pattern images. To prevent such an estimation error, we search and extract well-suited regions to estimate PSF in the projection image and estimate the PSF at extracted region. Experimental results show that our method can prevent PSF estimation error without any special images.Item Interactive ray shading of FRep objects(Václav Skala - UNION Agency, 2008) Fryazinov, Oleg; Pasko, Alexander; Cunningham, Steve; Skala, VáclavIn this paper we present a method for interactive rendering general procedurally defined functionally represented (FRep) objects using the acceleration with graphics hardware, namely Graphics Processing Units (GPU). We obtain interactive rates by using GPU acceleration for all computations in rendering algorithm, such as raysurface intersection, function evaluation and normal computations. We compute primary rays as well as secondary rays for shadows, reflection and refraction for obtaining high quality of the output visualization and further extension to ray-tracing of FRep objects. The algorithm is well-suited for modern GPUs and provides acceptable interactive rates with good quality of the results. A wide range of objects can be rendered including traditional skeletal implicit surfaces, constructive solids, and purely procedural objects such as 3D fractals.
