Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

OpenLiDARMap: Zero-Drift Point Cloud Mapping using Map Priors

Dominik Kulmer, Maximilian Leitenstern, Marcel Weinmann, Markus Lienkamp

TL;DR

OpenLiDARMap tackles GNSS-denied georeferenced mapping by fusing LiDAR data with sparse public map priors (building footprints and surface models) in a pose-graph framework that jointly performs scan-to-map and scan-to-scan registration. By tying absolute poses to a sparse reference map while preserving local consistency through incremental scan matching, it eliminates long-term drift without loop closures or learning-based components. The method demonstrates drift-free, georeferenced mapping across KITTI, NCLT, and EDGAR datasets and varying LiDAR setups, even with outdated priors, and runs efficiently (~30 ms per KITTI frame). An open-source implementation is provided to facilitate adoption and further research in GNSS-free mapping for robotics and autonomous systems.

Abstract

Accurate localization is a critical component of mobile autonomous systems, especially in Global Navigation Satellite Systems (GNSS)-denied environments where traditional methods fail. In such scenarios, environmental sensing is essential for reliable operation. However, approaches such as LiDAR odometry and Simultaneous Localization and Mapping (SLAM) suffer from drift over long distances, especially in the absence of loop closures. Map-based localization offers a robust alternative, but the challenge lies in creating and georeferencing maps without GNSS support. To address this issue, we propose a method for creating georeferenced maps without GNSS by using publicly available data, such as building footprints and surface models derived from sparse aerial scans. Our approach integrates these data with onboard LiDAR scans to produce dense, accurate, georeferenced 3D point cloud maps. By combining an Iterative Closest Point (ICP) scan-to-scan and scan-to-map matching strategy, we achieve high local consistency without suffering from long-term drift. Thus, we eliminate the reliance on GNSS for the creation of georeferenced maps. The results demonstrate that LiDAR-only mapping can produce accurate georeferenced point cloud maps when augmented with existing map priors.

OpenLiDARMap: Zero-Drift Point Cloud Mapping using Map Priors

TL;DR

OpenLiDARMap tackles GNSS-denied georeferenced mapping by fusing LiDAR data with sparse public map priors (building footprints and surface models) in a pose-graph framework that jointly performs scan-to-map and scan-to-scan registration. By tying absolute poses to a sparse reference map while preserving local consistency through incremental scan matching, it eliminates long-term drift without loop closures or learning-based components. The method demonstrates drift-free, georeferenced mapping across KITTI, NCLT, and EDGAR datasets and varying LiDAR setups, even with outdated priors, and runs efficiently (~30 ms per KITTI frame). An open-source implementation is provided to facilitate adoption and further research in GNSS-free mapping for robotics and autonomous systems.

Abstract

Accurate localization is a critical component of mobile autonomous systems, especially in Global Navigation Satellite Systems (GNSS)-denied environments where traditional methods fail. In such scenarios, environmental sensing is essential for reliable operation. However, approaches such as LiDAR odometry and Simultaneous Localization and Mapping (SLAM) suffer from drift over long distances, especially in the absence of loop closures. Map-based localization offers a robust alternative, but the challenge lies in creating and georeferencing maps without GNSS support. To address this issue, we propose a method for creating georeferenced maps without GNSS by using publicly available data, such as building footprints and surface models derived from sparse aerial scans. Our approach integrates these data with onboard LiDAR scans to produce dense, accurate, georeferenced 3D point cloud maps. By combining an Iterative Closest Point (ICP) scan-to-scan and scan-to-map matching strategy, we achieve high local consistency without suffering from long-term drift. Thus, we eliminate the reliance on GNSS for the creation of georeferenced maps. The results demonstrate that LiDAR-only mapping can produce accurate georeferenced point cloud maps when augmented with existing map priors.
Paper Structure (16 sections, 3 equations, 9 figures, 4 tables)

This paper contains 16 sections, 3 equations, 9 figures, 4 tables.

Table of Contents

  1. INTRODUCTION
  2. RELATED WORK
  3. GEOREFENCED POINT CLOUD MAPPING
  4. Generating Sparse Reference Maps
  5. Scan-to-Map Matching
  6. Scan-to-Scan Matching
  7. Pose-Graph Optimization
  8. Estimating the Initial Guess
  9. Parameters
  10. EXPERIMENTAL EVALUATION
  11. Experimental Setup
  12. Global Displacement Evaluation
  13. Local Consistency Evaluation
  14. Out-of-Date Reference Maps
  15. LIMITATIONS AND FUTURE WORK
  16. ...and 1 more sections

Figures (9)

  • Figure 1: Representation of the different map priors and formats from the building approximations on the left to the final georeferenced point cloud map on the right for KITTI Seq. 00.
  • Figure 2: Exemplary illustration of a combined sparse point cloud map of (green) approximated building data and (black) the surface model with a ground sampling distance of 1m around the starting position of KITTI Seq. 00.
  • Figure 3: Pipeline overview that shows our main steps to generate georeferenced point cloud maps. Starting with (1) the sparse reference map, which is used for (2) the scan-to-map matching. (3) shows the scan-to-scan matching. The two results are then optimized with (4) a pose-graph optimization, resulting in the final pose, which is then used for (5) the initial guess estimation for the next LiDAR frame.
  • Figure 4: Absolute Trajectory Error (ATE) of our approach for KITTI Seq. 00. (I) Point cloud created from the "ground truth" data. (II) Point cloud created with our approach.
  • Figure 5: Absolute Trajectory Error (ATE) of our approach for the EDGAR Campus dataset.
  • ...and 4 more figures