Skip to content

Las Vegas Reconstruction 2.0 (LVR2)

  • Software
  • Navigation and mapping
  • Library

A C++ library for building surface reconstructions from point cloud data, with classification of the found surfaces into predefined categories, aimed at robotics such as navigation and localization.

Made by Osnabrück University – Mobile Robotics, Germany (checked on github.com/uos (opens in a new tab)).

Source on GitHub: github.com/uos/lvr2 (opens in a new tab). Facts read at commit 05a197e (opens in a new tab).

Licenses

Code
BSD-3-Clause (opens in a new tab)

Read by Robotheca from LICENSE (opens in a new tab) (BSD-3-Clause) at commit 05a197e (opens in a new tab), on 16 Oct 2026.

Commercial use
Yes

What it does

  • Builds surface meshes from point clouds, reading inputs such as ASCII .pts/.xyz, .ply and .h5 files.
  • Estimates point normals using several methods, including PCA, RANSAC and IPCA.
  • Offers several voxel decomposition methods for reconstruction, such as Marching Cubes and Planar Marching Cubes.
  • Classifies the reconstructed surfaces into predefined categories and colors the mesh accordingly.
  • Can fill holes, remove small regions and dangling artifacts, and optimize planar regions.
  • Includes texture generation and texture matching tools, and can be used from a CMake project or ROS 2 workspace.

Written by Robotheca from the project's README (opens in a new tab) at commit 05a197e.

Getting started

Difficulty
Advanced
Hardware it needs, as the project states it
A C++17 compiler is required. Installation is documented for Ubuntu 18.04 to 24.04 and MacOS via Homebrew; CUDA support is optional for NVIDIA graphics cards.

Maintenance

Status
Active: a commit in the last 90 days
Newest commit
6 Aug 2026 on main
Latest release
25.2.3 (30 Mar 2026)

Checked on GitHub on 16 Oct 2026.

Robots that can run it

No robot on Robotheca is linked to it yet. Links come from a robot's robot.yaml, from its dependency files read as text, or from Robotheca's editors.

Robotheca lists software as a reference: nothing is hosted or run here, and every fact links to where it was read.