Parallel view-dependent refinement of progressive meshes

Abstract

We present a scheme for view-dependent level-of-detail control that is implemented entirely on programmable graphics hardware. Our scheme selectively refines and coarsens an arbitrary triangle mesh at the granularity of individual vertices, to create meshes that are highly adapted to dynamic view parameters. Such fine-grain control has previously been demonstrated using sequential CPU algorithms. However, these algorithms involve pointer-based structures with intricate dependencies that cannot be handled efficiently within the restricted framework of GPU parallelism. We show that by introducing new data structures and dependency rules, one can realize fine-grain progressive mesh updates as a sequence of parallel streaming passes over the mesh elements. A major design challenge is that the GPU processes stream elements in isolation. The mesh update algorithm has time complexity proportional to the selectively refined mesh, and moreover can be amortized across several frames. The static data structure is remarkably compact, requiring only 57% more memory than an indexed triangle list. We demonstrate real-time exploration of complex models with normals and textures.

Parallel View-Dependent Level-of-Detail Control

“…Our system is the first to perform real-time vertexgranular LOD over arbitrary triangle meshes on the GPU [26]. This work extends the presentation of our approach and provides additional usage scenarios for this level-of-detail control.…”

Integrating occlusion culling with parallel LOD for rendering complex 3D environments on GPU

“…The representative vertex position for each cluster can be independently computed through geometry shader stage. [Hu et al 2009] introduced a GPU-based approach for viewdependent Progressive Meshes, where vertex dependencies were not fully considered during a cascade of vertex splitting events. [Derzapf et al 2010] encoded the dependency information of Progressive Meshes into a GPU-friendly compact data structure.…”

“…However, the requirement to interactively render gigabyte-scale models usually overburdens the computational power and memory capacity of the GPUs. To solve this problem, GPU-based parallel mesh simplification algorithms, such as [Hu et al 2009;Derzapf et al 2010;Peng et al 2011]), have been proposed to fast simplify complex models. However, without considering view-parameters, the occluded objects are undesirably rendered in high levels of details, where processing them consumes many computational resources.…”

Simplification and streaming of GIS terrain for web clients

“…• The 1-ring of R is traversed to update the triangles by replacing R with P or C Currently, we face the same problem as Hu et al [Hu et al 2009]: in practice, the addition of points in the buffer of candidates is performed sequentially. However, as Hu et al, we remain confident that rapidly, the mechanisms provided by the GPU should resolve this particular problem.…”

“…The problem of mesh visualization performed by the GPU was also addressed by Hu et al [Hu et al 2009], but this work does not aim to compute new mesh versions but rather to visualize viewdependent intermediate meshes: the content of these intermediate meshes is computed in a preprocessing step, then the algorithm load them in the GPU memory.…”