Compression and rendering of voxelized 3D scene representations


Compression and rendering of voxelized 3D scene representations



With the increase in performance and programmability of graphical processing units (GPUs), GPU raycasting is emerging as an efficient solution for many real-time rendering problems in a variety of application domains, from industrial CAD rendering to gaming to cultural heritage data presentation.

In order to handle large detailed scenes, devising a compact and efficient scene representation for accelerating ray-geometry intersection queries becomes paramount and many solutions have been proposed.

Among these, sparse voxel octrees (SVO) have provided impressive results, since they can be created from a variety of scene representation, they efficiently carve out empty space, with benefits on ray tracing performance and memory needs and they implicitly provide a levels-of-detail (LOD) mechanism.

Given their still relatively high memory cost and the required high memory bandwidth, these voxelized approaches have, however, been limited to moderate scene sizes and resolutions, or to effects that do not require precise geometric details.

It is therefore important to devise novel scene representations that can provide compact representations within reasonable memory footprints, while not requiring decompression overhead.


CRS4 has created and developed methods for efficient lossless compression of voxelized geometry in a format that keeps the same visibility-query performance. Our approach profits from redundancy and symmetry present in common scenes by merging subtrees that are identical through a similarity transform. Moreover, we exploit the skewed distribution of references to shared nodes to store child pointers using a variable bit-rate encoding. We have also shown how, by selecting plane reflections along the main grid directions as symmetry transforms, we can construct highly compressed GPU-friendly structures using a fully out-of-core method. Our results demonstrate that state-of-the-art compression and real-time tracing performance can be achieved on high-resolution voxelized representations of real-world scenes of very different characteristics, including large CAD models, 3D scans, and typical gaming models. The method has been implemented in an open-source software that covers compression and rendering (SSVDAG).

Innovative features

  • state-of-the-art compression performance of sparse voxel octrees using advanced deduplication and pointer compression;
  • exploitation of symmetries in a GPU friendly structure;
  • complete solutions for scalable processing and rendering of voxelized representation starting from massive triangulated surface models.

Potential users

Researchers in visual computing, engineers active in industrial settings and gaming.

Impact sectors

ICT - Cultural Heritage - Gaming.

Other resources

  1. SSVDAG framework
  2. Alberto Jaspe Villanueva. Scalable Exploration of 3D Massive Models. PhD thesis. PhD Programme in Information and Communications Technology, University of A Coruña, Spain, 2018.
  3. Alberto Jaspe Villanueva, Fabio Marton, and Enrico Gobbetti. Symmetry-aware Sparse Voxel DAGs (SSVDAGs) for compression-domain tracing of high-resolution geometric scenes. Journal of Computer Graphics Techniques, 6(2): 1-30, 2017.
  4. Alberto Jaspe Villanueva, Fabio Marton, and Enrico Gobbetti. SSVDAGs: Symmetry-aware Sparse Voxel DAGs. In Proc. ACM i3D. Pages 7-14, February 2016.

Questo sito utilizza cookie tecnici e assimilati. Possono essere presenti anche cookie profilazione di terze parti. Se vuoi saperne di più o negare il consenso a tutti o ad alcuni cookie leggi l'informativa completa. Proseguendo nella navigazione (anche con il semplice scrolling) acconsenti all'uso dei cookie. This site uses technical and anonymized analytics cookies only. There may also be profiling third-party cookies. Please read the cookie information page to learn more about how we use cookies or blocking them. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.