Bevy Point Cloud

WASM Demo site : https://rlamarche.github.io/bevy_pointcloud/

A Bevy plugin to render point clouds.

Roadmap :

• Render point cloud using a shader
• Compute AABB
• Add helpers for importing las/laz files
• Add helpers for importing PLY files
• Implement rendering described in https://www.cg.tuwien.ac.at/research/publications/2016/SCHUETZ-2016-POT/SCHUETZ-2016-POT-thesis.pdf
• Read Potree files in https://www.cg.tuwien.ac.at/research/publications/2016/SCHUETZ-2016-POT/SCHUETZ-2016-POT-thesis.pdf
• Efficiently load/unload chunks based on camera location
• ...

Takanawa Lion with multi pass rendering pipeline inspired by Potree:

Takanawa Lion with multi pass rendering pipeline + Eye Dome Lighting inspired by Potree:

1 million random points:

St. Anna Mountain, 33 millions points:

Credits

• Rendering techniques and shaders adapted from Potree
  Copyright (c) 2011-2020, Markus Schütz
  Licensed under the BSD 2-Clause License (see THIRD_PARTY_LICENSES.md).

• Potree itself is based on prior research and open source projects.
  Notably:

  • Multi-resolution octree algorithms developed at the Vienna University of Technology (Michael Wimmer, Claus Scheiblauer, Scanopy Project).
  • Shaders inspired by CloudCompare (EDL), plas.io (LAS/LAZ support), and other point cloud tools.

• The rendering pipeline implementation is based on Bevy's render pipelines
  (Bevy Engine, MIT or Apache-2.0).

Acknowledgements

This project would not have been possible without the amazing work of the open-source and research communities. In particular, I would like to acknowledge:

• Potree by Markus Schütz, which inspired the rendering techniques and some shader code used here.
• Bevy Engine, whose rendering pipelines served as a foundation for the Rust-side implementation.
• The Vienna University of Technology (Michael Wimmer, Claus Scheiblauer) for their research on multi-resolution octrees in the Scanopy Project.
• CloudCompare for the EDL shader approach.
• plas.io for early contributions to LAS/LAZ point cloud rendering.
• The broader open-source graphics and visualization community for their continuous contributions to point-based rendering.

Thank you to all the researchers, developers, and contributors whose work laid the ground for this project.

Prepare WASM Example

Run the following commands to prepare the WASM:

cargo build --features webgl --example simple --target wasm32-unknown-unknown  --release
wasm-bindgen --no-typescript --target web  --out-dir path/to/site/ --out-name "bevy_pointcloud"  ./target/wasm32-unknown-unknown/release/examples/simple.wasm

Optimize the wasm:

• Note: use the latest version of https://github.com/WebAssembly/binaryen because the one shipped with Ubuntu 24.04 make a broken WASM.

wasm-opt -O -ol 100 -s 100 -o bevy_pointcloud/bevy_pointcloud_bg.wasm bevy_pointcloud/bevy_pointcloud_bg.wasm