Entity

Time filter

Source Type

Sardinia, Italy

Rodriguez M.B.,Visual Computing Group | Alcocer P.P.V.,Polytechnic University of Catalonia
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2012

Volume rendering has been a relevant topic in scientific visualization for the last two decades. A decade ago the exploration of reasonably big volume datasets required costly workstations due to the high processing cost of this kind of visualization. In the last years, a high end PC or laptop was enough to be able to handle medium-sized datasets thanks specially to the fast evolution of GPU hardware. New embedded CPUs that sport powerful graphics chipsets make complex 3D applications feasible in such devices. However, besides the much marketed presentations and all its hype, no real empirical data is usually available that makes comparing absolute and relative capabilities possible. In this paper we analyze current graphics hardware in most high-end Android mobile devices and perform a practical comparison of a well-known GPU-intensive task: volume rendering. We analyze different aspects by implementing three different classical algorithms and show how the current state-of-the art mobile GPUs behave in volume rendering. © 2012 Springer-Verlag. Source


Gunther T.,Visual Computing Group | Theisel H.,Otto Von Guericke University of Magdeburg
Computer Graphics Forum | Year: 2016

Inertial particles are finite-sized objects traveling with a certain velocity that differs from the underlying carrying flow, i.e., they are mass-dependent and subject to inertia. Their backward integration is in practice infeasible, since a slight change in the initial velocity causes extreme changes in the recovered position. Thus, if an inertial particle is observed, it is difficult to recover where it came from. This is known as the source inversion problem, which has many practical applications in recovering the source of airborne or waterborne pollutions. Inertial trajectories live in a higher dimensional spatio-velocity space. In this paper, we show that this space is only sparsely populated. Assuming that inertial particles are released with a given initial velocity (e.g., from rest), particles may reach a certain location only with a limited set of possible velocities. In fact, with increasing integration duration and dependent on the particle response time, inertial particles converge to a terminal velocity. We show that the set of initial positions that lead to the same location form a curve. We extract these curves by devising a derived vector field in which they appear as tangent curves. Most importantly, the derived vector field only involves forward integrated flow map gradients, which are much more stable to compute than backward trajectories. After extraction, we interactively visualize the curves in the domain and display the reached velocities using glyphs. In addition, we encode the rate of change of the terminal velocity along the curves, which gives a notion for the convergence to the terminal velocity. With this, we present the first solution to the source inversion problem that considers actual inertial trajectories. We apply the method to steady and unsteady flows in both 2D and 3D domains. © 2016 The Author(s) Computer Graphics Forum © 2016 The Eurographics Association and John Wiley & Sons Ltd. Published by John Wiley & Sons Ltd. Source


Mura C.,University of Zurich | Mattausch O.,University of Zurich | Villanueva A.J.,Visual Computing Group | Gobbetti E.,Visual Computing Group | Pajarola R.,University of Zurich
Proceedings - 13th International Conference on Computer-Aided Design and Computer Graphics, CAD/Graphics 2013 | Year: 2013

We present a robust approach for reconstructing the architectural structure of complex indoor environments given a set of cluttered input scans. Our method first uses an efficient occlusion-aware process to extract planar patches as candidate walls, separating them from clutter and coping with missing data. Using a diffusion process to further increase its robustness, our algorithm is able to reconstruct a clean architectural model from the candidate walls. To our knowledge, this is the first indoor reconstruction method which goes beyond a binary classification and automatically recognizes different rooms as separate components. We demonstrate the validity of our approach by testing it on both synthetic models and real-world 3D scans of indoor environments. © 2013 IEEE. Source


Mura C.,University of Zurich | Mattausch O.,University of Zurich | Jaspe Villanueva A.,Visual Computing Group | Gobbetti E.,Visual Computing Group | Pajarola R.,University of Zurich
Computers and Graphics (Pergamon) | Year: 2014

We present a robust approach for reconstructing the main architectural structure of complex indoor environments given a set of cluttered 3D input range scans. Our method uses an efficient occlusion-aware process to extract planar patches as candidate walls, separating them from clutter and coping with missing data, and automatically extracts the individual rooms that compose the environment by applying a diffusion process on the space partitioning induced by the candidate walls. This diffusion process, which has a natural interpretation in terms of heat propagation, makes our method robust to artifacts and other imperfections that occur in typical scanned data of interiors. For each room, our algorithm reconstructs an accurate polyhedral model by applying methods from robust statistics. We demonstrate the validity of our approach by evaluating it on both synthetic models and real-world 3D scans of indoor environments. © 2014 The Authors. Source


Zagoris K.,Greek National Center For Scientific Research | Zagoris K.,Visual Computing Group | Pratikakis I.,Visual Computing Group | Gatos B.,Greek National Center For Scientific Research
Proceedings of International Conference on Frontiers in Handwriting Recognition, ICFHR | Year: 2014

Many word spotting strategies for the modern documents are not directly applicable to historical handwritten documents due to writing styles variety and intense degradation. In this paper, a new method that permits effective word spotting in handwritten documents is presented that relies upon document-specific local features which take into account texture information around representative key points. Experimental work on two historical handwritten datasets using standard evaluation measures shows the improved performance achieved by the proposed methodology. © 2014 IEEE. Source

Discover hidden collaborations