# hku-mars/fast-livo2 **Attribution required: if you use, quote, or summarise this content, you must credit and link back to [awesome-repositories.com](https://awesome-repositories.com/repository/hku-mars-fast-livo2).** 3,634 stars · 627 forks · C++ · gpl-2.0 ## Links - GitHub: https://github.com/hku-mars/FAST-LIVO2 - awesome-repositories: https://awesome-repositories.com/repository/hku-mars-fast-livo2.md ## Topics `3d-reconstruction` `colored-point-cloud` `gaussian-splatting` `lidar-camera-fusion` `lidar-inertial-odometry` `lidar-slam` `mesh-reconstruction` `nerf` `sensor-fusion` `slam` ## Description FAST-LIVO2 is a LiDAR-inertial odometry framework and factor-graph SLAM implementation designed for real-time robot localization and 3D mapping. It functions as a multi-sensor fusion pipeline and state estimator that integrates LiDAR, inertial, and camera inputs to track a robot's position and orientation. The system employs a tightly-coupled sensor fusion approach to maintain stable navigation, particularly in degraded environments. It utilizes a voxel-based 3D mapping tool to organize point clouds into volumetric grids, which optimizes memory usage and search speed during spatial reconstruction. The framework covers a broad range of capabilities including autonomous robot localization, real-time 3D environment reconstruction, and odometry estimation. These are supported by a processing architecture that incorporates factor graph optimization, Kalman filter state estimation, and iterative closest point matching to resolve sensor drift and contradictions. ## Tags ### Hardware & IoT - [LiDAR-Inertial Odometry Frameworks](https://awesome-repositories.com/f/hardware-iot/lidar-inertial-odometry-frameworks.md) — Provides a complete framework for real-time robot localization and 3D mapping using tightly-coupled LiDAR and inertial sensor fusion. - [Robotics And Autonomous Systems](https://awesome-repositories.com/f/hardware-iot/embedded-robotics/robotics-autonomous-systems.md) — Determines robot coordinates within a map to enable autonomous navigation and movement. - [Sensor Fusion](https://awesome-repositories.com/f/hardware-iot/embedded-robotics/robotics-autonomous-systems/localization-mapping/sensor-fusion.md) — Combines various sensor inputs using probabilistic filtering to maintain accurate localization in challenging environments. ([source](https://github.com/hku-mars/FAST-LIVO2#readme)) - [Multi-Modal SLAM Implementations](https://awesome-repositories.com/f/hardware-iot/embedded-robotics/robotics-autonomous-systems/localization-mapping/slam-algorithms/visual-slam-implementations/multi-modal-slam-implementations.md) — Implements a multi-modal SLAM approach using factor graph optimization to reduce drift and sensor contradictions. - [State Estimation Libraries](https://awesome-repositories.com/f/hardware-iot/embedded-robotics/robotics-autonomous-systems/localization-mapping/state-estimation-libraries.md) — Tracks robot movement and orientation by combining multiple onboard sensor data streams. ([source](https://github.com/hku-mars/FAST-LIVO2/custom-properties)) ### Software Engineering & Architecture - [Tightly-Coupled LiDAR-Inertial Fusion](https://awesome-repositories.com/f/software-engineering-architecture/lidar-mapping/tightly-coupled-lidar-inertial-fusion.md) — Provides a tightly-coupled LiDAR-inertial fusion pipeline for high-frequency robot motion tracking. - [Kalman Filter Localization](https://awesome-repositories.com/f/software-engineering-architecture/kalman-filter-localization.md) — Employs Kalman filter state estimation to predict robot position and orientation using inertial data. - [Lidar Mapping](https://awesome-repositories.com/f/software-engineering-architecture/lidar-mapping.md) — Generates real-time 3D reconstructions by combining light, inertial, and camera data. ([source](https://github.com/hku-mars/FAST-LIVO2/tree/main/launch)) ### Artificial Intelligence & ML - [Sensor Fusion](https://awesome-repositories.com/f/artificial-intelligence-ml/sensor-fusion.md) — Integrates raw LiDAR and inertial measurements into a single state estimator for reliable navigation. ### Part of an Awesome List - [Real-Time Environmental Reconstruction](https://awesome-repositories.com/f/awesome-lists/ai/3d-reconstruction/real-time-environmental-reconstruction.md) — Builds detailed three-dimensional maps of the surrounding area in real-time while maintaining spatial stability. ([source](https://github.com/hku-mars/FAST-LIVO2#readme)) - [Factor Graphs](https://awesome-repositories.com/f/awesome-lists/ai/graph-representation-learning/factor-graphs.md) — Uses factor graphs to resolve sensor drift and optimize a network of spatial constraints. - [Optimization Algorithms](https://awesome-repositories.com/f/awesome-lists/ai/graph-representation-learning/factor-graphs/optimization-algorithms.md) — Implements factor-graph-based optimization to resolve sensor contradictions and reduce drift in robot localization. - [SLAM and Odometry](https://awesome-repositories.com/f/awesome-lists/ai/slam-and-odometry.md) — Tracks precise robot position and orientation by fusing high-frequency inertial measurements with laser scan data. - [Voxel-Based Grids](https://awesome-repositories.com/f/awesome-lists/ai/point-cloud-and-3d-processing/3d-point-cloud-representations/voxel-based-grids.md) — Utilizes a voxel-based mapping tool to organize point clouds into volumetric grids, optimizing memory usage and search speed. - [Spatial Mapping](https://awesome-repositories.com/f/awesome-lists/data/spatial-mapping.md) — Implements voxel-based spatial mapping to organize point clouds into volumetric grids for efficient search. ### Data & Databases - [Real-Time 3D Occupancy Mappers](https://awesome-repositories.com/f/data-databases/data-processing-pipelines/data-transformation/array-tensor-manipulation/array-filtering/grid-generation/occupancy-grid-generators/real-time-3d-occupancy-mappers.md) — Builds detailed 3D reconstructions of physical environments in real-time as a robot moves. ### Graphics & Multimedia - [Sparse Voxel Representations](https://awesome-repositories.com/f/graphics-multimedia/sparse-voxel-representations.md) — Organizes 3D spatial data into volumetric grids using sparse voxel representations to optimize memory. ### Scientific & Mathematical Computing - [Iterative Closest Point Matching](https://awesome-repositories.com/f/scientific-mathematical-computing/iterative-closest-point-matching.md) — Uses iterative closest point matching to align current LiDAR scans to a global map.