Figure 1: Four different views of a scene where agents navigate on the surface of a complex triangular mesh. Agents are color-coded by their different objectives. The system supports path planning of multiple agents on non-planar surfaces, and imposes no limitation on the domain mesh, such as this mesh with more than one genus. AbstractPath planning is an active topic in the literature, and efficient navigation over non-planar surfaces is an open research question. In this work we present a novel technique for navigation of multiple agents over arbitrary triangular domains. The proposed solution uses a fast hierarchical computation of geodesic distances over triangular meshes to allow interactive frame rates, and a GPU-based collision avoidance technique to guide individual agents. Unlike most previous work, the method imposes no limitations on the surface over which the agents are moving, and can naturally deal with non-planar meshes of arbitrary genus and curvature. Moreover, the implementation is a hybrid CPU/GPU algorithm that explores the current trend of increasing the number of CPU cores and GPU programmability. This approach exploits the best qualities in each processor, thus achieving very high performance.

Interactive isosurface extraction has recently become possible through successful efforts to map algorithms such as Marching Cubes (MC) and Marching Tetrahedra (MT) to modern Graphics Processing Unit (GPU) architectures. Other isosurfacing algorithms, however, are not so easily portable to GPUs, either because they involve more complex operations or because they are not based on discrete case tables, as is the case with most marching techniques. In this paper, we revisit the Dual Contouring (MC) and Macet isosurface extraction algorithms and propose, respectively: (i) a novel, efficient and parallelizable version of Dual Contouring and (ii) a set of GPU modules which extend the original Marching Cubes algorithm. Similar to marching methods, our novel technique is based on a case table, which allows for a very efficient GPU implementation. In addition, we enumerate and evaluate several alternatives to implement efficient contouring algorithms on the GPU, and present trade-offs among all approaches. Finally, we validate the efficiency and quality of the tessellations produced in all these alternatives.

No abstract

Fig. 1:Merging a buddha tetrahedral mesh with a background grid. Our technique is able to handle meshes with distinct levels of refinement -observe how the internal tetrahedra of the buddha have not been refined.Abstract-Simplicial meshes are extremely useful as discrete approximations of continuous spaces in numerical simulations. In some applications, however, meshes need to be modified over time. Mesh update operations are often expensive and brittle, which tends to make the numerical simulations unstable. In this paper we propose an alternative technique for updating simplicial meshes that undergo geometric and topological changes. We exploit the property that a Weighted Delaunay Triangulation (WDT) can be used to implicitly define the connectivity of a mesh. Instead of explicitly maintaining connectivity information, we simply keep a collection of weights associated with each vertex. This approach allows for a simple way to merge triangulations, which we illustrate with examples in 2D and 3D.

While there exist popular software tools that leverage the power of arrays of tiled high resolution displays, they usually require either the use of a particular API or significant programming effort to be properly configured. We present PVW (Parallel Visualization using display Walls), a framework that uses display walls for scientific visualization, requiring minimum labor in setup, programming and configuration. PVW works as a plug-in to pipeline-based visualization software, and allows users to migrate existing visualizations designed for a singleworkstation, single-display setup to a large tiled display running on a distributed machine. Our framework is also extensible, allowing different APIs and algorithms to be made display wall-aware with minimum effort.