📡 Embodied AI Radar具身智能雷达

Topics主题 › Perception, SLAM & 3D Mapping感知、SLAM 与三维建图

Paper论文 2021 · HKU MaRS Lab

FAST-LIO2: Fast Direct LiDAR-inertial Odometry

Registers raw LiDAR points directly to the map with an incremental k-d tree; fast enough for onboard drone and robot computers.用增量 k-d 树把原始激光点直接配准到地图,速度足以在无人机和机器人机载计算机上运行。

BibTeX
@misc{xu2021fast,
  title = {{FAST-LIO2: Fast Direct LiDAR-inertial Odometry}},
  author = {Wei Xu and Yixi Cai and Dongjiao He and Jiarong Lin and Fu Zhang},
  year = {2021},
  eprint = {2107.06829},
  archivePrefix = {arXiv},
  primaryClass = {cs.RO},
  url = {https://arxiv.org/abs/2107.06829},
}
Perception, SLAM & 3D Mapping感知、SLAM 与三维建图Drone / aerial无人机lidarodometryreal-time

Authors (from arXiv): 作者(来自 arXiv):Wei Xu, Yixi Cai, Dongjiao He, Jiarong Lin, Fu Zhang

More from: 同机构更多条目:The University of Hong Kong香港大学

More in Perception, SLAM & 3D Mapping更多感知、SLAM 与三维建图条目

All Perception, SLAM & 3D Mapping entries →查看全部感知、SLAM 与三维建图条目 →