Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

No More Potentially Dynamic Objects: Static Point Cloud Map Generation based on 3D Object Detection and Ground Projection

Soojin Woo, Donghwi Jung, Seong-Woo Kim

TL;DR

This work addresses the problem of dynamic objects in LiDAR point clouds degrading map quality and localization. It introduces a pipeline that detects dynamic objects in a single frame using a voxel-based 3D detector and then projects the dynamic points onto the ground via ground segmentation, producing a static point cloud map. The method combines 3D voxel-based detection (VoxelNeXt), ground segmentation (Patchwork++), and LOAM-based SLAM with loop closure to maintain map consistency while eliminating both currently dynamic and potentially dynamic objects. Evaluations on KITTI and campus datasets show preserved mapping performance and improved localization when using the generated static maps, and the authors publicly release their code. This approach offers a practical, timing-invariant static-map generation technique for robust autonomous driving localization and planning.

Abstract

In this paper, we propose an algorithm to generate a static point cloud map based on LiDAR point cloud data. Our proposed pipeline detects dynamic objects using 3D object detectors and projects points of dynamic objects onto the ground. Typically, point cloud data acquired in real-time serves as a snapshot of the surrounding areas containing both static objects and dynamic objects. The static objects include buildings and trees, otherwise, the dynamic objects contain objects such as parked cars that change their position over time. Removing dynamic objects from the point cloud map is crucial as they can degrade the quality and localization accuracy of the map. To address this issue, in this paper, we propose an algorithm that creates a map only consisting of static objects. We apply a 3D object detection algorithm to the point cloud data which are obtained from LiDAR to implement our pipeline. We then stack the points to create the map after performing ground segmentation and projection. As a result, not only we can eliminate currently dynamic objects at the time of map generation but also potentially dynamic objects such as parked vehicles. We validate the performance of our method using two kinds of datasets collected on real roads: KITTI and our dataset. The result demonstrates the capability of our proposal to create an accurate static map excluding dynamic objects from input point clouds. Also, we verified the improved performance of localization using a generated map based on our method.

No More Potentially Dynamic Objects: Static Point Cloud Map Generation based on 3D Object Detection and Ground Projection

TL;DR

This work addresses the problem of dynamic objects in LiDAR point clouds degrading map quality and localization. It introduces a pipeline that detects dynamic objects in a single frame using a voxel-based 3D detector and then projects the dynamic points onto the ground via ground segmentation, producing a static point cloud map. The method combines 3D voxel-based detection (VoxelNeXt), ground segmentation (Patchwork++), and LOAM-based SLAM with loop closure to maintain map consistency while eliminating both currently dynamic and potentially dynamic objects. Evaluations on KITTI and campus datasets show preserved mapping performance and improved localization when using the generated static maps, and the authors publicly release their code. This approach offers a practical, timing-invariant static-map generation technique for robust autonomous driving localization and planning.

Abstract

In this paper, we propose an algorithm to generate a static point cloud map based on LiDAR point cloud data. Our proposed pipeline detects dynamic objects using 3D object detectors and projects points of dynamic objects onto the ground. Typically, point cloud data acquired in real-time serves as a snapshot of the surrounding areas containing both static objects and dynamic objects. The static objects include buildings and trees, otherwise, the dynamic objects contain objects such as parked cars that change their position over time. Removing dynamic objects from the point cloud map is crucial as they can degrade the quality and localization accuracy of the map. To address this issue, in this paper, we propose an algorithm that creates a map only consisting of static objects. We apply a 3D object detection algorithm to the point cloud data which are obtained from LiDAR to implement our pipeline. We then stack the points to create the map after performing ground segmentation and projection. As a result, not only we can eliminate currently dynamic objects at the time of map generation but also potentially dynamic objects such as parked vehicles. We validate the performance of our method using two kinds of datasets collected on real roads: KITTI and our dataset. The result demonstrates the capability of our proposal to create an accurate static map excluding dynamic objects from input point clouds. Also, we verified the improved performance of localization using a generated map based on our method.
Paper Structure (13 sections, 3 equations, 7 figures, 2 tables)

This paper contains 13 sections, 3 equations, 7 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Methods
  4. 3D Object Detection
  5. Ground Segmentation
  6. Projection and Mapping
  7. Experiments
  8. Implementation
  9. Data
  10. Evaluation
  11. Mapping
  12. Localization
  13. Conclusion

Figures (7)

  • Figure 1: Changed distribution of potentially dynamic objects over time at the same location.
  • Figure 2: The process of our static map creation method.
  • Figure 3: The process of voxel-based 3D object detection.
  • Figure 4: The projection result of the point cloud onto the ground plane within the 3D bounding box obtained by the 3D object detection model. Dynamic objects (pink), Static objects (gray).
  • Figure 5: Comparison of SLAM odometry prediction performance on the 00 sequence of KITTI dataset.
  • ...and 2 more figures