PRBonn/MapClosures

Effectively Detecting Loop Closures using Point Cloud Density Maps

What it solves

MapClosures addresses the challenge of effectively detecting loop closures in LiDAR-based SLAM (Simultaneous Localization and Mapping). Loop closure detection is critical for correcting drift in a robot's estimated trajectory by recognizing when the robot has returned to a previously visited location.

How it works

The project utilizes point cloud density maps to identify and match previously visited areas, allowing the system to recognize loop closures more efficiently and accurately than traditional methods.

Who it’s for

This tool is designed for robotics researchers and developers working with LiDAR-based SLAM systems who need a reliable way to handle loop closure detection.

Highlights

  • Provides both a C++ library for high-performance integration and a Python API/CLI for ease of use.
  • Based on research published in ICRA 2024 and IJRR 2026.
  • Includes a visualizer to inspect the loop closure pipeline.
  • Built upon and inspired by the KISS-ICP framework.

Related

  • Project
    leggedrobotics/open3d_slam

    A C++ library for pointcloud-based SLAM designed for education, enabling users to build maps or localize within them using basic, well-established algorithms.

  • Project
    introlab/rtabmap_ros

    A ROS 2 wrapper for RTAB-Map that provides graph-based SLAM with appearance-based loop closure detection for building 3D maps from RGB-D, stereo, or lidar data.

  • Project
    rsasaki0109/lidar_localization_ros2

    A map-based 3D LiDAR localization system for ROS 2 and Nav2 that uses NDT and GICP to align real-time scans with prior point-cloud maps.

  • Project
    AlanLuSun/High-quality-ellipse-detection

    A high-quality ellipse detection algorithm for real-world images, offering high localization accuracy and efficiency for applications like robot vision and PCB inspection.

  • Project
    neka-nat/probreg

    A library for point cloud registration using probabilistic models, offering a more robust alternative to ICP for both rigid and non-rigid transformations.