Time filter

Source Type

Schmid J.,University of Geneva | Iglesias Guitian J.A.,CRS4 Visual Computing Group | Gobbetti E.,CRS4 Visual Computing Group | Magnenat-Thalmann N.,University of Geneva | Magnenat-Thalmann N.,Nanyang Technological University
Visual Computer | Year: 2011

Despite the ability of current GPU processors to treat heavy parallel computation tasks, its use for solving medical image segmentation problems is still not fully exploited and remains challenging. A lot of difficulties may arise related to, for example, the different image modalities, noise and artifacts of source images, or the shape and appearance variability of the structures to segment. Motivated by practical problems of image segmentation in the medical field, we present in this paper a GPU framework based on explicit discrete deformable models, implemented over the NVidia CUDA architecture, aimed for the segmentation of volumetric images. The framework supports the segmentation in parallel of different volumetric structures as well as interaction during the segmentation process and real-time visualization of the intermediate results. Promising results in terms of accuracy and speed on a real segmentation experiment have demonstrated the usability of the system. © 2010 Springer-Verlag.

Marton F.,CRS4 Visual Computing Group | Gobbetti E.,CRS4 Visual Computing Group | Bettio F.,CRS4 Visual Computing Group | Iglesias Guitian J.A.,CRS4 Visual Computing Group | Pintus R.,CRS4 Visual Computing Group
3DTV Conference: The True Vision - Capture, Transmission and Display of 3D Video, 3DTV-CON 2011 - Proceedings | Year: 2011

We present an end-to-end system capable of real-time capturing and displaying with full horizontal parallax high-quality 3D video contents on a cluster-driven multiprojector light-field display. The capture component is an array of low-cost USB cameras connected to a single PC. RawM-JPEG data coming fromthe software-synchronized cameras are multicast over Gigabit Ethernet to the back-end nodes of the rendering cluster, where they are decompressed and rendered. For all-in-focus rendering, view-dependent depth is estimated on the GPU using a customized multiview space-sweeping approach based on fast Census-based area matching implemented in CUDA. Realtime performance is demonstrated on a system with 18 VGA cameras and 72 SVGA rendering projectors. © 2011 IEEE.

Pintore G.,CRS4 Visual Computing Group | Gobbetti E.,CRS4 | Ganovelli F.,CNR Institute of Neuroscience | Brivio P.,CNR Institute of Neuroscience
Proceedings, Web3D 2012 - 17th International Conference on 3D Web Technology | Year: 2012

We report on the 3DNSITE system, a web-based client-server 3D visualization tool for streaming and visualizing large tridimensional hybrid data (georeferenced point clouds and photographs with associated viewpoints and camera parameters). The system is motivated by the need to simplify data acquisition and location recognition for crisis managers and first responders during emergency operations or training sessions. In this peculiar context, it is very important to easily share 3D environment data among people in a distributed environment, accessing huge 3D models with embedded photographs on devices with heterogenous hardware capabilities and interconnected on different network types. Moreover, since the specific end-users are not necessary skilled with virtual reality and 3D objects interaction, the navigation interface must be simple and intuitive. Taking into account these constraints, we propose a mixel object-based/image-based system, which enhances the current state-of-the-art by exploiting a multi-resolution representation for the 3D model and a multi-level cache system for both the images and 3D models structure. A novel low-degree-of-freedom user interface is presented to navigate in the scenario with touchscreen devices. The proposed implementation, included in a more general training and decision framework for emergency operations, is evaluated on real-world datasets. © 2012 ACM.

Loading CRS4 Visual Computing Group collaborators
Loading CRS4 Visual Computing Group collaborators