introlab/rtabmap

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

3965 stars · 954 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

RTAB-Map (Real-Time Appearance-Based Mapping) is a C++ library and standalone application implementing graph-based SLAM for RGB-D, stereo, and lidar sensors, with an appearance-based loop closure detector. It supports 6DoF and 3DoF mapping on robots or handheld devices, with ROS/ROS2 integration and mobile support.

Use cases

• build 3D maps of indoor environments with an RGB-D camera or lidar
• perform real-time SLAM and loop closure detection on a robot
• localize a robot against a previously built map
• do multi-session mapping with illumination-invariant re-localization
• create 6DoF scans with a handheld Kinect or stereo camera
• integrate SLAM into a ROS or ROS2 robotics stack

When to choose

• you need a mature, actively maintained open-source SLAM solution supporting RGB-D, stereo, and lidar
• you want ROS/ROS2 integration for a robotics project
• you need large-scale, long-term online mapping with memory management
• you want multi-session mapping and re-localization out of the box

When to avoid

• you need visual-inertial odometry only without mapping or loop closure
• you want a lightweight pure-visual odometry pipeline without a graph backend
• your platform is not among Linux, macOS, Windows, Android, or iOS

Facets

library · maturity stable

computer-vision simulation sdk cli robotics autonomous-vehicles computer-vision windows cpp python cross-platform slam mapping localization rgb-d lidar stereo-vision loop-closure ros ros2 graph-optimization 3d-mapping algorithms linux macos android ios

Member repositories

For agents

markdown · JSON · MCP: product_card(name="introlab/rtabmap")