Abstract: The advent of GPGPU technologies has allowed for sensible speed-ups in many high-dimension, memory-intensive computational problems. In this paper we demonstrate the effectiveness of such techniques by describing two applications of GPGPU computing to two different subfields of computer graphics, namely computer vision and mesh processing. In the first case, CUDA technology is employed to accelerate the computation of approximation of motion between two images, known also as optical flow. As for mesh processing, we exploit the massively parallel architecture of CUDA devices to accelerate the face clustering procedure that is employed in many recent mesh segmentation algorithms. In both cases, the results obtained so far are presented and thoroughly discussed, along with the expected future development of the work.
Authors: S. Marras, C. Mura, E. Gobbetti, R. Scateni, R. Scopigno.
Two examples of GPGPU acceleration of memory-intensive algorithm.
EuroGraphics Italian Chapter 2010, 49-56.
Genova, Italia, Novembre 2010.
Abstract: Wall-sized interactive displays gain more and more attention as a valuable tool for multiuser applications, but typically require the adoption of projectors tiles. Projectors tend to display deformed images, due to lens distortion and/or imperfection, and because they are almost never perfectly aligned to the projection surface. Multi-projector videowalls are typically bounded to the video architecture and to the specific application to be displayed. This makes it harder to develop interactive applications, in which a fine grained control of the coordinate transformations (to and from user space and model space) is required. This paper presents a solution to such issues: implementing the blending functionalities at an application level allows seamless development of multi-display interactive applications with multi-touch capabilities. The description of the multi-touch interaction, guaranteed by an array of cameras on the baseline of the wall, is beyond the scope of this work which focuses on calibration.
Abstract: In this paper we propose an optimization of the Shape Diameter Function (SDF) that we call Accelerated SDF (ASDF). We discuss in detail the advantages and disadvantages of the original SDF definition, proposing theoretical and practical approaches for speedup and approximation. Using Poisson-based interpolation we compute the SDF value for a small subset of randomly distributed faces and propagate the values over the mesh. We show the results obtained with ASDF versus SDF in terms of timings and error.
Abstract: Merging meshes is a recurrent need in geometry modeling and it is a critical step in the 3D acquisition pipeline, where it is used for building a single mesh from several range scans. A pioneering simple and effective solution to merging is represented by the Zippering algorithm (Turk and Levoy, 1994), which consists of simply stitching the meshes together along their borders. In this paper we propose a new extended version of the zippering algorithm that enables the user to control the resulting mesh by introducing quality criteria in the selection of redundant data, and allows to zip together meshes with different granularity by an ad hoc refinement algorithm.
Abstract: The Education Programme at Eurographics 2009 took place in Munich, Germany, over the course of 2 days: March 31st and April 1st 2009. Educators were invited to present their experience in teaching computer graphics over a wide range of topics: from teaching mathematical foundations of computer graphics to using visual tools; from teaching in a strict computer science curriculum to teaching students of mixed disciplines and on to teaching in other curricula. As a result, we had 12 presentations in four sessions, ranging from a new method of teaching quaternions to teaching computer graphics in the context of theatre. The presence of 20–40 attendees throughout these 2 days made clear that the Education Programme at Eurographics has established itself over the last years.