Efficient and Probabilistic Adaptive Voxel Mapping for Accurate Online LiDAR Odometry

Yuan, Chongjian; Wei, Xu; Liu, Xiyuan; Hong, Xiaoping; Zhang, Fu

doi:10.1109/lra.2022.3187250

Cited by 70 publications

(35 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this section, we extensively evaluate the performance of ImMesh. Notice that our localization module is built upon our previous work VoxelMap [16] with no modification that relative to the state estimation. Hence, the localization precision of this work performs as well as [16].…”

Section: Experiments and Resultsmentioning

confidence: 99%

“…The localization module utilizes the input data stream of receiver module, real-time estimating the sensor poses of 6 DoF and registering the points to map. Our localization module is built upon our previous work VoxelMap [16], which represents the surrounding environment with the probabilistic representation, estimating pose with an iterated Kalman filter by maximum a posterior.…”

Section: Receiver and Localizationmentioning

confidence: 99%

“…We only discuss those processes in localization module that are closely related to our meshing module in this paper. For the detailed modeling and analysis of LiDAR's measurement noise, we recommend our readers to our previous work VoxelMap [16].…”

Section: A Voxel Map Constructionmentioning

confidence: 99%

“…1) Update of voxel map: The registered LiDAR points are used for constructing the probabilistic voxel map by updating the point distributions (i.e, A in ( 12)), planar parameters (i.e., n, q) and the correspondent uncertainties of all possible hierarchical voxels. For the details of this voxel map update, we refer the reader to our previous work [16]. Besides, if a new register point does not lie on an existing L2 (or L1) voxel, a new L2 (or L1) voxel will be created and added to the hash table, after, this point will be added to the newly constructed voxel.…”

Section: Point Cloud Registrationmentioning

confidence: 99%

“…‚ We propose a novel system that can estimate the sensor pose and reconstruct the mesh of the surrounding environment both online. Its localization is built upon our previous work VoxelMap [16], which can estimate the sensor pose of better efficiency and higher accuracy over its counterparts (e.g., FAST-LIO2 [17], SUMA [18], MULLS [19], Lego-LOAM [20], etc.). Its meshing module implements a novel mesh reconstruction approach, which efficiently reconstructs the mesh in an incremental manner, and can achieve real-time performance in largescale scenarios on a standard desktop CPU.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

ImMesh: An Immediate LiDAR Localization and Meshing Framework

Lin¹,

Yuan²,

Cai³

et al. 2023

Preprint

View full text Add to dashboard Cite

with pull, commit and push steps for incremental reconstruction of triangle facets. To the best of our knowledge, this is the first work in literature that can reconstruct online the triangle mesh of large-scale scenes, just relying on a standard CPU without GPU acceleration. To share our findings and make contributions to the community, we make our code publicly available on our GitHub: github.com/hku-mars/ImMesh.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

Section: Receiver and Localizationmentioning

confidence: 99%

Section: A Voxel Map Constructionmentioning

confidence: 99%

Section: Point Cloud Registrationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

ImMesh: An Immediate LiDAR Localization and Meshing Framework

Lin¹,

Yuan²,

Cai³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

Point‐LIO: Robust High‐Bandwidth Light Detection and Ranging Inertial Odometry

Wei

Chen

et al. 2023

Advanced Intelligent Systems

View full text Add to dashboard Cite

Herein, point light detection and ranging inertial odometry (LIO) is presented: a robust and high‐bandwidth light detection and ranging (LiDAR) inertial odometry with the capability to estimate extremely aggressive robotic motions. Point‐LIO has two key novelties. The first one is a point‐by‐point LIO framework that updates the state at each LiDAR point measurement. This framework allows an extremely high‐frequency odometry output, significantly increases the odometry bandwidth, and fundamentally removes the artificial in‐frame motion distortion. The second one is a stochastic process‐augmented kinematic model which models the IMU measurement as an output. This new modeling method enables accurate localization and reliable mapping for aggressive motions even with inertial measurement unit (IMU) measurements saturated in the middle of the motion. Various real‐world experiments are conducted for performance evaluation. Overall, Point‐LIO is capable to provide accurate, high‐frequency odometry (4–8 kHz) and reliable mapping under severe vibrations and aggressive motions with high angular velocity (75 rad s−1) beyond the IMU measuring ranges. Furthermore, an exhaustive benchmark comparison is conducted. Point‐LIO achieves consistently comparable accuracy and time consumption. Finally, two example applications of Point‐LIO are demonstrated, one is a racing drone and the other is a self‐rotating unmanned aerial vehicle, both have aggressive motions.

show abstract