Baike.dev
All toolsOnline toolsAI codingTrendingOpen sourceNewsSubmit
Log in
< Back to tools
M

MS-Mapping

> 编程语言
Open source

[ICRA@40] MS-Mapping: An Uncertainty-Aware Large-Scale Multi-Session LiDAR Mapping System

378 stars0 likes0 views
WebsiteGitHub

About

[ICRA@40] MS-Mapping: An Uncertainty-Aware Large-Scale Multi-Session LiDAR Mapping System

MS-Mapping is a multi-session LiDAR mapping system designed for large-scale environments. It addresses challenges in data redundancy, robustness, and accuracy with three key innovations:

  • Distribution-aware keyframe selection: Captures the contributions of each point cloud frame by analyzing map distribution similarities. This reduces data redundancy and optimizes graph size and speed.

  • Uncertainty model: Automatically adjusts using the covariance matrix during graph optimization, enhancing precision and robustness without scene-specific tuning. It monitors pose uncertainty to avoid ill-posed optimizations.

  • Enhanced evaluation: Redesigned baseline comparisons and benchmarks demonstrate MS-Mapping's superior accuracy over state-of-the-art methods.

News

  • 2025/05/16: Add docker support which adapted to Ubuntu 24.04 by @bboyack. Also add more databag with accurate GT trajectory and map in MS-dataset (Google Drive).
  • 2025/03/26: Add new databag RB3 and new Tutorial ! Feel free to pull issues for any questions related to this work!
  • 2025/02/26: Baseline methods F2F and M2F released! Tutorial is here!
  • 2024/08/08: We released the first version of MS-Mapping on ArXiv, together with the example merged data and related YouTube and bilibili videos.
  • 2024/07/19: accepted by ICRA@40 as a extended abstract.
  • 2024/06/03: submit to a workshop.

Tutorial Here !!!!

Dataset

Fusion Portable V2 Dataset Newer College Urban-Nav MS-Dataset

Download Test Data in Ms-Datset (Google Drive)

GT Map PK01 CP05 RB02 RB03 CS01
Merged results for 8 sessions CC01

Trajectory Evaluation

Map Evaluation

we use MapEval for this map evaluation.

Time Analysis

To plot the results, you can follow this scripts.

Quick Run

Install

  • Ubuntu 20.04 / ROS Noetic
  • Open3d ( >= 0.17.0)
  • PCL
  • GTSAM 4.2.0
  • CMake > 3.20 (fixed by @bboyack)
bash
# for cmake update, required by open3d 0.17.0
cd cmake-
./configure
make -j8
sudo make install
cmake --version

Docker Support by @bboyack

  • Ubuntu 24.04

Baselines

The implementation of baseline method F2F and M2F, only radius keyframe selection + fix-cov PGO. Tutorial is here!

  • Step 1: using old session with single session mode (useMultiMode = false) to prepare data (e.g., PK1).

  • Step2: incrimental mapping using new session rosbag (e.g., RB2).

  • Step3: global map merging with giving initial pose (manually from ClouCompare or place recognition methods), e.g. PK1-RB2.

PK1 PK1-RB2 PK1-RB2-RB3
  • Step4: Lifelong Mapping with BeautyMap to remove dynamic points.
Clean Map Using BeautyMap Ground Truth Map

TO DO

  • Clean codes
  • Add more dataset support
  • Add place recognition algothem for initialization
  • Add GNSS support. (users can refer to LIO_SAM_6AXIS to merge this codes)

Citations

Please cite:

…

Acknowledgment

The code in this project is adapted from the following projects:

  • The odometry method is adapted from FAST-LIO2.
  • The basic framework for pose graph optimization (PGO) is adapted from SC-A-LOAM.

Contributors

Issues· 0 open

View all issuesOpen on GitHub

No open issues yet, or sync has not completed.

> Tags

C++gtsamkeyframelarge-scale-mappinglidar

No comments yet. Be the first to share.

> Details

PublishedAug 1, 2026
UpdatedSep 17, 2026
Category编程语言
PricingOpen source

> Related tools

T
TypeScript
JavaScript 的超集,为前端与全栈提供静态类型
P
Python
通用编程语言,广泛用于 Web、数据与 AI
G
Go
Google 推出的简洁高效系统语言