hwan0806/Awesome-LiDAR-Mapping

What it solves

Robots that operate for long periods need accurate maps of their surroundings built from LiDAR sensors. But choosing the right mapping method is hard: there are many research papers, each with different strengths, and not all of them release code. This project solves that by gathering peer-reviewed LiDAR mapping papers that come with open-source implementations into one place, saving researchers and engineers from hunting through conference proceedings and GitHub on their own.

How it works

It is a curated list, organized by topic. The main sections cover LiDAR-only mapping, LiDAR-inertial fusion (adding motion sensors), multiple LiDAR setups, learning-based approaches, static map building (removing moving objects), and merging multiple maps. Each entry lists the publication year, venue, paper title, and a direct link to the code repository. The project also includes a section pointing to general SLAM frameworks and evaluation tools. It is updated periodically, as shown by the news log, and accepts contributions through pull requests or issues.

Who it’s for

Robotics researchers, graduate students, and engineers working on autonomous vehicles, drones, or any platform that builds 3D maps from LiDAR. It helps anyone who wants to find a ready-to-use mapping algorithm with open code rather than implementing one from scratch.

Highlights

• Covers 25+ LiDAR-only methods, from classic LOAM to recent real-time graph-based approaches.
• Dedicated section for LiDAR-inertial odometry, including FAST-LIO, LIO-SAM, and newer filter-based estimators.
• Includes multi-LiDAR calibration and mapping systems, plus learning-based mapping with neural representations.
• Static mapping section lists algorithms and learning models for removing dynamic objects, essential for long-term autonomy.
• Each paper links directly to its open-source repository with star counts, making quality easy to gauge.