This is a standalone project to make viewing massive point clouds easy and convenient. It was build to serve the needs of the Cartographer project, but is useful in its own right.

The project consist of a root crate that can build and read octrees on disk and viewer binaries to visualize the data. For Mac OS X, we assume below that you've installed Homebrew.

• Install Rust: curl https://sh.rustup.rs -sSf | sh. See https://rustup.rs for details.
• Install the rust gRPC protobuf plugin: cargo install grpcio-compiler and make sure it is in your $PATH. This has more dependencies, it requires cmake and go. On mac: brew install cmake go.
• Initialize all submodules: git submodule update --init --recursive.

In the root of the repo, run cargo build --release. Then use target/release/build_octree to generate an octree out of a PLY file.

This is a native client using SDL2.

1. Install SDL2. For example, on Mac brew install sdl2.
2. Change to the sdl viewer's directory: cd sdl_viewer/.
3. Build with cargo build --release.
4. Run with ../target/release/sdl_viewer <octree directory>.

In the point cloud viewer, navigate with the keyboard or with the mouse or touchpad. Dragging while pressing the left mouse button rotates, dragging while pressing the right mouse button pans the view. The following keys are bound:

Saved camera positions are persisted in the octree directory and will therefore live through restarts of the program.

The octree_web_viewer consists of TypeScript code running in the browser and a web server binary.

To build and run the octree_web_viewer please look into the octree_web_viewer README file

This work was inspired by the following projects.

• Potree
• Megatree