LIO-GVM: an accurate, tightly-coupled Lidar-inertial odometry with Gaussian voxel map
This letter presents a probabilistic voxel-based LiDAR Inertial Odometry framework for accurate and robust pose estimation. The framework addresses the correspondence mismatching issue by representing the LiDAR points as a set of Gaussian distributions and evaluating the divergence in variance for o...
| Main Authors: | , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/178008 |
| _version_ | 1826112902721961984 |
|---|---|
| author | Ji, Xingyu Yuan, Shenghai Yin, Pengyu Xie, Lihua |
| author2 | Centre for Advanced Robotics Technology Innovation (CARTIN) |
| author_facet | Centre for Advanced Robotics Technology Innovation (CARTIN) Ji, Xingyu Yuan, Shenghai Yin, Pengyu Xie, Lihua |
| author_sort | Ji, Xingyu |
| collection | NTU |
| description | This letter presents a probabilistic voxel-based LiDAR Inertial Odometry framework for accurate and robust pose estimation. The framework addresses the correspondence mismatching issue by representing the LiDAR points as a set of Gaussian distributions and evaluating the divergence in variance for outlier rejection. Based on the fitted distributions, a new residual metric is proposed for the filter-based Lidar inertial odometry by incorporating both the distance and variance disparities, further enriching the comprehensiveness and accuracy of the residual metric. With the strategic design of the residual, we propose a simple yet effective voxel-solely mapping scheme, which only requires the maintenance of one centroid and one covariance matrix for each voxel. Experiments on different datasets demonstrate the robustness and high accuracy of our framework for various data inputs and environments. |
| first_indexed | 2024-10-01T03:14:45Z |
| format | Journal Article |
| id | ntu-10356/178008 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T03:14:45Z |
| publishDate | 2024 |
| record_format | dspace |
| spelling | ntu-10356/1780082024-06-04T02:01:36Z LIO-GVM: an accurate, tightly-coupled Lidar-inertial odometry with Gaussian voxel map Ji, Xingyu Yuan, Shenghai Yin, Pengyu Xie, Lihua Centre for Advanced Robotics Technology Innovation (CARTIN) Engineering LiDAR inertial odometry Voxel map. This letter presents a probabilistic voxel-based LiDAR Inertial Odometry framework for accurate and robust pose estimation. The framework addresses the correspondence mismatching issue by representing the LiDAR points as a set of Gaussian distributions and evaluating the divergence in variance for outlier rejection. Based on the fitted distributions, a new residual metric is proposed for the filter-based Lidar inertial odometry by incorporating both the distance and variance disparities, further enriching the comprehensiveness and accuracy of the residual metric. With the strategic design of the residual, we propose a simple yet effective voxel-solely mapping scheme, which only requires the maintenance of one centroid and one covariance matrix for each voxel. Experiments on different datasets demonstrate the robustness and high accuracy of our framework for various data inputs and environments. National Research Foundation (NRF) This work was supported by the National Research Foundation, Singapore under its Medium Sized Center for Advanced Robotics Technology Innovation. 2024-06-04T02:01:35Z 2024-06-04T02:01:35Z 2024 Journal Article Ji, X., Yuan, S., Yin, P. & Xie, L. (2024). LIO-GVM: an accurate, tightly-coupled Lidar-inertial odometry with Gaussian voxel map. IEEE Robotics and Automation Letters, 9(3), 2200-2207. https://dx.doi.org/10.1109/LRA.2024.3354616 2377-3766 https://hdl.handle.net/10356/178008 10.1109/LRA.2024.3354616 2-s2.0-85182936586 3 9 2200 2207 en IEEE Robotics and Automation Letters © 2024 IEEE. All rights reserved. |
| spellingShingle | Engineering LiDAR inertial odometry Voxel map. Ji, Xingyu Yuan, Shenghai Yin, Pengyu Xie, Lihua LIO-GVM: an accurate, tightly-coupled Lidar-inertial odometry with Gaussian voxel map |
| title | LIO-GVM: an accurate, tightly-coupled Lidar-inertial odometry with Gaussian voxel map |
| title_full | LIO-GVM: an accurate, tightly-coupled Lidar-inertial odometry with Gaussian voxel map |
| title_fullStr | LIO-GVM: an accurate, tightly-coupled Lidar-inertial odometry with Gaussian voxel map |
| title_full_unstemmed | LIO-GVM: an accurate, tightly-coupled Lidar-inertial odometry with Gaussian voxel map |
| title_short | LIO-GVM: an accurate, tightly-coupled Lidar-inertial odometry with Gaussian voxel map |
| title_sort | lio gvm an accurate tightly coupled lidar inertial odometry with gaussian voxel map |
| topic | Engineering LiDAR inertial odometry Voxel map. |
| url | https://hdl.handle.net/10356/178008 |
| work_keys_str_mv | AT jixingyu liogvmanaccuratetightlycoupledlidarinertialodometrywithgaussianvoxelmap AT yuanshenghai liogvmanaccuratetightlycoupledlidarinertialodometrywithgaussianvoxelmap AT yinpengyu liogvmanaccuratetightlycoupledlidarinertialodometrywithgaussianvoxelmap AT xielihua liogvmanaccuratetightlycoupledlidarinertialodometrywithgaussianvoxelmap |