QSplat: a multiresolution point rendering system for large meshes

By Szymon Rusinkiewicz, Marc Levoy. 2000

1,250 citations recorded by the source corpus.

Advances in 3D scanning technologies have enabled the practical creation of meshes with hundreds of millions of polygons. Traditional algorithms for display, simplification, and progressive transmission of meshes are impractical for data sets of this size. We describe a system for representing and progressively displaying these meshes that combines a multiresolution hierarchy based on bounding spheres with a rendering system based on points. A single data structure is used for view frustum culling, backface culling, level-of-detail selection, and rendering. The representation is compact and can be computed quickly, making it suitable for large data sets. Our implementation, written for use in a large-scale 3D digitization project, launches quickly, maintains a user-settable interactive frame rate regardless of object complexity or camera position, yields reasonable image quality during motion, and refines progressively when idle to a high final image quality. We have demonstrated the system on scanned models containing hundreds of millions of samples.

Subfields: Computational Mechanics, Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Computer Graphics and Computer-Aided Design

In Computer Vision and Pattern Recognition

Citations in the corpus

Pointshop 3D: an interactive system for point-based surface editing (2002, background): “It is rather designed to complement low cost scanning devices [Eyetronics 2001] and point-based 3D viewers [Rusinkiewicz and Levoy 2000, Pfister et al. 2000, Arius3D 2001], yielding a powerful pipeline for efficient 3D content creation and display.”
The 3D Model Acquisition Pipeline (2002, methodology): “To make the results usable, they developed a novel rendering system that generates images directly from points rather than from triangle primitives [116].”
Visualization of large terrains made easy (2001, methodology): “In particular, we exploit the nested bounding sphere hierarchy to perform view culling, similar to [24].”
Perspective shadow maps (2002, background): “, [8, 10, 11, 14]) has proven to be a very effective means for rendering complex geometry.”
Point set surfaces (2001, methodology): “Rusinkiewicz and Levoy [36] introduce a technique which uses a hierarchy of spheres of different radii to model a high-resolution model.”
Building a Digital Model of Michelangelo's Florentine Pietà (2002, background): “Incorporating ideas such as pointbased rendering [15] or multipass texturing now available on inexpensive graphics cards would improve this aspect of our system.”
Estimating surface normals in noisy point cloud data (2003, background): “For certain applications like rendering or visualization, such a computation is often unnecessary and direct rendering of PCD has been investigated by the graphics community [14, 16].”
NeRF: Representing scenes as neural radiance ﬁelds for view synthesis (neutral): “In the past year, further breakthroughs have been achieved through an even older graphics technique of point-based rendering, where points are used as a display primitive [ 32, 57 ].”
Direct visibility of point sets (2007, methodology): “Since both rays and points are singular primitives, this requires the algorithm to either trace “thick” rays [Schaufler and Jensen 2000] or use “finite-area” points [Rusinkiewicz and Levoy 2000; Dutré et al. 2000; Zwicker et al. 2001; Wu and Kobbelt 2004; Guennebaud et al. 2004; Guennebaud et al. 20”
Computing and Rendering Point Set Surfaces (2003, methodology): “Another reason is that highly detailed surfaces require a large number of small primitives, which contribute to less than a pixel when displayed, so that points become an effective display primitive [41], [46].”
Surface splatting (2001, neutral): “QSplat [16] is a point rendering system that was designed to interactively render large data sets produced by modern scanning devices.”
Recent advances in 3D Gaussian splatting (2024, positive): “Later, a series of splatting methods aim to enhance it by introducing a texture filter for anti-aliasing rendering [8], improving rendering efficiency [9, 10], and resolving discontinuous shading [11].”