Culling an Object Hierarchy to a Frustum Hierarchy

Abstract

Abstract. Visibility culling of a scene is a crucial stage for interactive graphics applications, particularly for scenes with thousands of objects. The culling time must be small for it to be effective. A hierarchical representation of the scene is used for efficient culling tests. However, when there are multiple view frustums (as in a tiled display wall), visibility culling time becomes substantial and cannot be hidden by pipelining it with other stages of rendering. In this paper, we address the problem of culling an object to a hierarchically organized set of frustums, such as those found in tiled displays and shadow volume computation. We present an adaptive algorithm to unfold the twin hierarchies at every stage in the culling procedure. Our algorithm computes from-point visibility and is conservative. The precomputation required is minimal, allowing our approach to be applied for dynamic scenes as well. We show performance of our technique over different variants of culling a scene to multiple frustums. We also show results for dynamic scenes.

BqR‐Tree: A Data Structure for Flights and Walkthroughs in Urban Scenes with Mobile Elements

“…View frustum culling is an acceleration technique well studied in static environments [SMM08, Pla05, AM00] but it is rarely implemented in dynamic scenes. Seron et al [SRCP02] proposed two new types nodes for the integration of mobile elements into the scene graph, Nirnimesh et al [NHN06] proposed multiple view‐frustum.…”

A Survey of Software Frameworks for Cluster-Based Large High-Resolution Displays

“…Garuda relies on a server-push and multicast approach to handle the distributed scene graphs and dynamic 3D models across multiple display nodes. Each display node performs view frustum culling using a novel adaptive algorithm [68]. The framework also supports a non-invasive programming interface which automatically replaces OSG's cull, draw, and swap functions.…”

“…Since every node runs the same application, the na€ ıve approach would be for each node to attempt to render the entire scene, relying purely on the viewport clipping to reduce the amount of geometric information actually rendered and displayed. Thus, this model can be improved through support for object culling in the framework [68], or by intelligent processing at the application level to avoid considering objects that will not be displayed by this node. Of course, this later approach puts a greater burden on the application developer, and there is the potential that This model has some downsides in terms of utilization of the computational resources, performance scalability, and data management.…”