RainerKuemmerle/g2o

g2o: A General Framework for Graph Optimization

What it solves

g2o is a C++ framework designed to solve nonlinear least squares problems that can be represented as a graph. It is specifically used to find the optimal configuration of parameters or state variables that best explain a set of measurements affected by Gaussian noise, which is a common challenge in robotics and computer vision.

How it works

The framework optimizes graph-based nonlinear error functions. It is designed to be highly extensible, allowing users to specify new problems in just a few lines of code. The library provides built-in solutions for common robotics problems like Simultaneous Localization and Mapping (SLAM) and Bundle Adjustment (BA).

Who it’s for

It is intended for developers and researchers in robotics and computer vision who need a high-performance framework for graph-based optimization and state estimation.

Highlights

  • Extensible Architecture: New optimization problems can be defined quickly with minimal code.
  • High Performance: Offers performance comparable to state-of-the-art implementations for SLAM and BA.
  • Broad Compatibility: Supports Linux, Mac OS X, Windows, and Android.
  • C++17 Standard: Built using modern C++ standards and integrates with Eigen3.
  • Python Support: Experimental Python wrappers are available via the pymem branch and external bindings.

Related

  • Project
    borglab/gtsam

    A C++ library for smoothing and mapping in robotics and vision that uses factor graphs and Bayes networks for efficient state estimation and optimization.

  • Project
    MOLAorg/mola

    A modular C++ and ROS 2 framework for robotics localization and SLAM, enabling configurable pipelines for LiDAR odometry and high-precision mapping.

  • Project
    luigifreda/pyslam

    A hybrid Python/C++ modular framework for Visual SLAM that integrates modern AI depth prediction, semantic segmentation, and volumetric reconstruction for research and prototyping.

  • Project
    koide3/glim

    A versatile 3D mapping framework that uses factor graphs and GPU acceleration to provide accurate, sensor-agnostic range-based mapping and localization.

  • Project
    mlpack/ensmallen

    A high-quality C++ header-only library providing a wide range of non-linear numerical optimizers, including gradient-based and gradient-free methods.