Number 1-3 (2008)
Permanent URI for this collection
Browse
Recent Submissions
Item Robust face recognition under varying light based on 3D recovery(Václav Skala - UNION Agency, 2008) Yang, Guan; Xue, Fei; Feng, GuoCan; Suen, Ching Y.; Skala, VáclavThis paper addresses face recognition under varying light via 3D reconstruction based on the techniques of shape from shading (SFS). First, we improve the geometric-based SFS by introducing the integrability constraint as one of the regular terms. This operation preserves the local curvedness of the recovered surface. Second, we propose a novel method to investigate human face recognition in the illumination varying case using local topographic information, such as curvedness and shape index extracted from intensity images by SFS algorithms. The experimental results have shown that the curvedness and shape index are suitable for representing 3D local features, and also it is insensitive to light variations since only 3D information is involved. Compared with typical face recognition approaches based on principal component analysis (PCA) plus linear discriminant analysis (LDA), the proposed method has demonstrated a better performance. This implies local topological properties are effective attributes for face recognition under light variations
Item MPIglut: powerwall programming made easier(Václav Skala - UNION Agency, 2008) Lawlor, Orion Sky; Page, Matthew; Genetti, Jon; Skala, VáclavA powerwall is an array of separate screens that work together to provide a single unified display. Powerwalls are often driven by a small cluster, which requires parallel software to organize and synchronize the distributed rendering process. This paper describes MPIglut, our powerwall-friendly implementation of the popular sequential GLUT OpenGL 3D programming interface. MPIglut internally communicates using MPI to provide a single coherent display even across a distributedmemory parallel machine. Uniquely, MPIglut is sourcecode compatible with ordinary sequential GLUT code while providing high performance.Item Progressive combined B-reps – multi-resolution meshes for interactive real-time shape design(Václav Skala - UNION Agency, 2008) Havemann, Sven; Fellner, Dieter W.; Skala, VáclavWe present the Combined B-rep (cB-rep) as a multiresolution data structure suitable for interactive modeling and visualization of models composed of both free-form and polygonal parts. It is based on a half-edge data structure combined with Catmull/Clark subdivision surfaces. In addition to displaying the curved parts of the surface at an adaptive level-of-detail, the control mesh itself can be changed interactively at runtime using Euler operators. The tessellation of changed parts of the mesh is incrementally updated in real time. All changes in the mesh are logged, so that a complete undo/redo mechanism can be provided. We introduce Euler macros as a grouping mechanism for Euler operator sequences. The macro dependency graph, a directed acyclic graph, can be used for creating progressively increasing resolutions of the control mesh, and to guide the view-dependent refinement (pcB-rep). We consider Progessive Combined B-reps to be of use for data visualization and interactive 3D modeling, as well as a compact representation of synthetic 3D models.Item Moving with the flow: wave particles in flowing liquids(Václav Skala - UNION Agency, 2008) Cords, Hilko; Skala, VáclavThis paper presents a real-time water simulation framework with respect to wave propagation in flowing liquids. The presented method is based on the coupling of liquid flow simulation and simulation of surface waves. Physically, our system combines the solution of the Navier-Stokes Equations with the solution of the 2D Wave Equation. Numerically, we combine the concept of wave particles with a FDM-based flow simulation. Thus, wave propagation in fast flowing liquids (e.g., a creek) can be simulated – in real-time. Therefore, it is very suitable for today’s video games and VR-environments. Finally, we discuss the coupling of the liquid simulation with a rigid-body simulation and a fountain simulation.Item GPU bucket sort algorithm with applications to nearest-neighbour search(Václav Skala - UNION Agency, 2008) Roźen, T.; Boryczko, K.; Alda, Witold; Skala, VáclavWe present an adoption of the bucket sort algorithm capable of running entirely on GPU architecture. Our implementation employs render-to-texture to enable scatter operation. Linked lists of elements in each bucket are build and stored directly in video memory. We show also the use of this sorting method in a particle-based simulation. Dissipative Particle Dynamics is the physical model of choice; the simulation is performed entirely on the graphics hardware. GPU bucket sorting is used to build nearest-neighbour maps on a regular cell-grid which are the input of interparticle interaction computation. Finally we implement a simple random-number generator which is required by the DPD method.Item Managing dynamic entities in mobile, urban virtual environments(Václav Skala - UNION Agency, 2008) Nurminen, Antti; Skala, VáclavMobile networked virtual environments (mNVE’s) are a new, emerging type of virtual environments. Mobile 3D maps that support dynamic entities and communication between clients are a subcategory of mNVE’s, intended for navigation and locationbased information browsing. Models and entities portrayed in 3D maps represent real environments and entities, such as buildings, vehicles and people. Our main contribution is in developing a lightweight and scalable scheme for real dynamic entity management and visibility culling by exploiting geometry of urban environments, the honesty of locally positioned clients and the lack of interference between clients. We bind moving entities to a topological network consisting of street segments, crossings and larger areas, all associated to precalculated visibility cells. Our system reduces visibility determination to a simple cell occupation logic, performed at smart clients or proxies. In this scheme, servers act as fast message passing switches, managing client subscription and query tables, simply forwarding state update messages. Computational scalability is ensured by transferring computations to client side, and networking scalability by spatially localized servers, which allow roaming by subscribing to each others’ neighboring visibility cells.Item Libugrab: a versatile grabbing library(Václav Skala - UNION Agency, 2008) Kahlesz, Ferenc; Klein, Reinhard; Skala, VáclavCurrent commercial and freely available grabbing libraries are tightly coupled to operating systems and/or imaging hardware. Moreover, they usually do not support any kind of distributed camera-systems. This forces developers to either reimplement significant parts of the application or to come up with elaborate abstraction for the grabbing, should the underlying operating system, hardware (e.g. changing from analog PAL sources to IIDC cameras) or distribution model (e.g. adding remote intelligent cameras, which are capable of image processing themselves) change. In this paper we describe ‘libugrab’, a versatile grabbing library designed to provide a flexible abstraction of the grabbing process. The main advantages of ‘libugrab’ over similar libraries are the following: open source license, cross-platform availability, network transparency, support for both push and pull grabbing models, built-in support for image-processing via callbacks. The design especially facilitates rapid prototyping of distributed vision systems, which we demonstrate by several examples.Item Automatic camera pose initialization, using scale, rotation and luminance invariant natural feature tracking(Václav Skala - UNION Agency, 2008) Bastos, Rafael; Dias, Miguel Sales; Skala, VáclavThe solution to the camera registration and tracking problem serves Augmented Reality, in order to provide an enhancement to the user’s cognitive perception of the real world and his/her situational awareness. By analyzing the five most representative tracking and feature detection techniques, we have concluded that the Camera Pose Initialization (CPI) problem, a relevant sub-problem in the overall camera tracking problem, is still far from being solved using straightforward and non-intrusive methods. The assessed techniques often use user inputs (i.e. mouse clicking) or auxiliary artifacts (i.e. fiducial markers) to solve the CPI problem. This paper presents a novel approach to real-time scale, rotation and luminance invariant natural feature tracking, in order to solve the CPI problem using totally automatic procedures. The technique is applicable for the case of planar objects with arbitrary topologies and natural textures, and can be used in Augmented Reality. We also present a heuristic method for feature clustering, which has revealed to be efficient and reliable. The presented work uses this novel feature detection technique as a baseline for a real-time and robust planar texture tracking algorithm, which combines optical flow, backprojection and template matching techniques. The paper presents also performance and precision results of the proposed technique.Item Interactive cutting operations for generating anatomical illustrations from volumetric data sets(Václav Skala - UNION Agency, 2008) Mensmann, Jörg; Ropinski, Timo; Hinrichs, Klaus; Skala, VáclavIn anatomical illustrations deformation is often used to increase expressivity, to improve spatial comprehension and to enable an unobstructed view onto otherwise occluded structures. Based on our analysis and classification of deformations frequently found in anatomical textbooks we introduce a technique for interactively creating such deformations of volumetric data acquired with medical scanners. Our approach exploits the 3D ChainMail algorithm in combination with a GPU-based ray-casting renderer in order to perform deformations. Thus complex, interactive deformations become possible without a costly preprocessing or the necessity to reduce the data set resolution. For cutting operations we provide a template-based interaction technique which supports precise control of the cutting parameters. For commonly used deformation operations we provide adaptable interaction templates, whereas arbitrary deformations can be specified by using a point-and-drag interface.Item GPU radiosity for triangular meshes with support of normal mapping and arbitrary light distributions(Václav Skala - UNION Agency, 2008) Wallner, Günter; Skala, VáclavThis paper describes an implementation of a progressive radiosity algorithm for triangular meshes which works completely on programmable graphics processors. Errors due to the rasterization of triangles are fixed in a post-processing step or with a fragment shader during runtime. Adaptive subdivision to increase the accuracy of the radiosity solution can be performed during render-time. Since we found that the gradient is not very robust to determine whether triangles should be subdivided or not, we propose a new technique which uses hardware occlusion queries to determine shadow boundaries in image space. The GPU implementation facilitates the simple integration of normal mapping into the radiosity process. Light distribution textures (LDTs) enable us to simulate a variety of real world light sources without much computational overhead. The derivation of such an LDT from a EULUMDAT file is described.Item Stable integration of the dynamic cosserat equations with application to hair modeling(Václav Skala - UNION Agency, 2008) Sobottka, Gerrit; Lay, Tomás; Weber, Andreas; Skala, VáclavIn this paper we propose a new method for stable numerical integration of the dynamic Cosserat equations for rods, which constitute a mechanical framework for the physically based modeling of slender structures like DNA strands, drill strings, marine cables or human hair. Our integration method is well-established in the field of structural dynamics and has the major advantage of unconditional stability as well as user controllable numerical damping. We demonstrate its advantages in the context of fiber-based modeling of human hair. To our knowledge this approach has not been used in the computer graphics community before.Item Scale space based feature point detection on surfaces(Václav Skala - UNION Agency, 2008) Schlattmann, Markus; Degener, Patrick; Klein, Reinhard; Skala, VáclavThe detection of stable feature points is an important preprocessing step for many applications in computer graphics. Especially, registration and matching often require feature points and depend heavily on their quality. In the 2D image case, scale space based feature detection is well established and shows unquestionably good results. We introduce a novel scale space generalization to 3D embedded surfaces for extracting surface features. In contrast to a straightforward generalization to 3D images our approach extracts intrinsic features. We argue that such features are superior, in particular in the context of partial matching. Our features are robust to noise and provide a good description of the object’s salient regions.Item Comparing GPLVM approaches for dimensionality reduction in character animation(Václav Skala - UNION Agency, 2008) Quirion, Sébastien; Duchesne, Chantale; Laurendeau, Denis; Marchand, Mario; Skala, VáclavGaussian Process Latent Variable Models (GPLVMs) have been found to allow dramatic dimensionality reduction in character animations, often yielding two-dimensional or three-dimensional spaces from which the animation can be retrieved without perceptible alterations. Recently, many researchers have used this approach and improved on it for their purposes, thus creating a number of GPLVM-based approaches. The current paper introduces the main concepts behind GPLVMs and introduces its most widely known variants. Each approach is then compared based on various criteria pertaining to the task of dimensionality reduction in character animation. In the light of our experiments, no single approach is preferred over all others in all respects. Depending whether dimensionality reduction is used for compression purposes, to interpolate new natural looking poses or to synthesize entirely new motions, different approaches will be preferred.Item Constrained shepard method for modeling and visualization of scattered data(Václav Skala - UNION Agency, 2008) Mustafa, Ghulan; Shah, Abad Ali; Asim, Muhammad Rafiq; Skala, VáclavThe problem of modelling and visualization of scattered data where there are inherit constraints on value of the data exist in many scientific and business research areas. For example, the value of mass concentration always has the lower bound of 0 and upper bound of 1. The modeling functions having gradient usually do not guarantee to preserve the bounds of data. In this paper we present the Constrained Shepard method for interpolation of scattered data satisfying the lower and upper bounds specified by the two constraint functions. The constrained interpolant is an extension of the Modified Quadratic Shepard method with comparable efficiency and accuracy. The proposed method is easy to implement and extend to higher dimensionality. The constrained interpolation function is Cd continuus.Item Fast incident light field acquisition and rendering(Václav Skala - UNION Agency, 2008) Ihrke, Ivo; Stich, Timo; Gottschlich, Heiko; Magnor, Marcus; Seidel, Hans-Peter; Skala, VáclavWe present a practical and inexpensive approach for the acquisition and rendering of static incident light fields. Incident light fields can be used for lighting virtual scenes or to insert virtual objects into real world video footage. The virtual objects are correctly lit and cast shadows in the same way as real objects in the scene. We propose to use an inexpensive planar mirror and a high dynamic range video camera to record incident light fields quickly, making our method suitable for outdoor use. The mirror serves as a moving virtual camera sampling the light field. To render with the acquired data we propose a hardware accelerated rendering algorithm that reproduces the complex lighting and shadows of the 4D light field. The algorithm is scalable and allows continuous trade between quality and rendering speed.Item Presampled visibility for ambient occlusion(Václav Skala - UNION Agency, 2008) Gaitatzes, Athanasios; Chrysanthou, Yiorgos; Papaioannou, Georgios; Skala, VáclavWe present a novel method to accelerate the computation of the visibility function of the lighting equation, in dynamic scenes composed of rigid, non-penetrating objects. The main idea of the technique is to pre-compute for each object in the scene its associated four-dimensional field that describes the visibility in each direction for all positional samples on a sphere around the object, we call this a displacement field. We are able to speed up the calculation of algorithms that trace visibility rays to near real time frame rates. The storage requirements of the technique, amounts from one byte to one bit per ray direction making it particularly attractive to scenes with multiple instances of the same object, as the same cached data can be reused, regardless of the geometric transformation applied to each instance. We suggest an acceleration technique and identify the sampling method that gives the best results based on experimentation.Item Real-time capable system for hand gesture recognition Using hidden Markov models in stereo color image sequences(Václav Skala - UNION Agency, 2008) Elmezain, Mahmoud; Al-Hamadi, Ayoub; Michaelis, Bernd; Skala, VáclavThis paper proposes a system to recognize the alphabets and numbers in real time from color image sequences by the motion trajectory of a single hand using Hidden Markov Models (HMM). Our system is based on three main stages; automatic segmentation and preprocessing of the hand regions, feature extraction and classification. In automatic segmentation and preprocessing stage, YCbCr color space and depth information are used to detect hands and face in connection with morphological operation where Gaussian Mixture Model (GMM) is used for computing the skin probability. After the hand is detected and the centroid point of the hand region is determined, the tracking will take place in the further steps to determine the hand motion trajectory by using a search area around the hand region. In the feature extraction stage, the orientation is determined between two consecutive points from hand motion trajectory and then it is quantized to give a discrete vector that is used as input to HMM. The final stage so-called classification, Baum-Welch algorithm (BW) is used to do a full train for HMM parameters. The gesture of alphabets and numbers is recognized by using Left-Right Banded model (LRB) in conjunction with Forward algorithm. In our experiment, 720 trained gestures are used for training and also 360 tested gestures for testing. Our system recognizes the alphabets from A to Z and numbers from 0 to 9 and achieves an average recognition rate of 94.72%.Item Parallel visualization of the sloan digital sky survey DR6(Václav Skala - UNION Agency, 2008) Domonkos, Balász; Ralovich, Kristóf; Skala, VáclavThe new generation astronomy digital archives cover large area of the sky at fine resolution in many wavelengths from ultraviolet through optical and infrared. For instance, one of these projects the Sloan Digital Sky Survey is creating a detailed catalog covering more than a quarter of the sky with images measured with five different filters. The size of the data set can be measured in terabytes. These archives enable astronomers to explore the data for their research. However, virtually walking through these huge data sets also enables to visualize the beauty of the Universe and raises problems which can be interesting for people related to computer graphics. In this paper we present a technique for parallel visualization of large-scale scattered astrophysical data that has wide-spectrum photometric property. Our method performs sort-last parallel particle rendering using hierarchical, static data distribution; and its performance scales up linearly by increasing the number of the rendering nodes. It also enables setting the color matching function in the rendering phase and as well as altering the distance calculation formulae that calculates spatial coordinates from the redshift – all interactively.Item vSLRcam – taking pictures in virtual environments(Václav Skala - UNION Agency, 2008) Brennecke, Angela; Panzer, Christian; Schlechtweg, Stefan; Skala, VáclavOur work presents a virtual single lens reflection camera (vSLRcam) application which is employed in a virtual training environment for crime scene investigation. vSLRcam’s back-end is a GPU based simulation of a realistic camera model taking into account SLR camera properties like apperture, shutter speed, lens, etc., as well as their interdependencies. Thus, we can obtain realistic lens effects like motion blur or depth of field in real-time. The application user interface allows for parameterizing the inidividual camera attributes to achieve those effects and, as a result, to take realistic pictures of the scene. The resulting images come very close to real world photographs with equal parameter values. Our main contributions are a common framework for the SLR camera attributes and the simulation of their interdependecies in a single application which is capable of rendering photographic lens effects in real-time.Item BCC-splines: generalization of B-splines for the body-centered cubic lattice(Václav Skala - UNION Agency, 2008) Csébfalvi, Balázs; Skala, VáclavRecently, the B-spline family of reconstruction filters has been generalized for the hexagonal lattice, which is optimal for sampling 2D circularly band-limited signals. In this paper, we extend this generalization to the body-centered cubic (BCC) lattice, which is optimal for sampling spherically band-limited 3D signals. We call the obtained new reconstruction filters BCC-splines. Although the explicit analytical formulas are not defined yet, we evaluate the discrete approximation of these filters in the frequency domain in order to analyze their performance in a volume-rendering application. Our experimental results show that the BCC-splines can be superior over the box splines previously proposed for the BCC lattice.
