RKHS-BA: A Robust Correspondence-Free Multi-View Registration Framework with Semantic Point Clouds
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915041278689280 |
|---|---|
| author | Zhang, Ray Song, Jingwei Gao, Xiang Wu, Junzhe Liu, Tianyi Zhang, Jinyuan Eustice, Ryan Ghaffari, Maani |
| author_facet | Zhang, Ray Song, Jingwei Gao, Xiang Wu, Junzhe Liu, Tianyi Zhang, Jinyuan Eustice, Ryan Ghaffari, Maani |
| contents | This work reports a novel multi-frame Bundle Adjustment (BA) framework called RKHS-BA. It uses continuous landmark representations that encode RGB-D/LiDAR and semantic observations in a Reproducing Kernel Hilbert Space (RKHS). With a correspondence-free pose graph formulation, the proposed system constructs a loss function that achieves more generalized convergence than classical point-wise convergence. We demonstrate its applications in multi-view point cloud registration, sliding-window odometry, and global LiDAR mapping on simulated and real data. It shows highly robust pose estimations in extremely noisy scenes and exhibits strong generalization with various types of semantic inputs. The open source implementation is released in https://github.com/UMich-CURLY/RKHS_BA. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_01254 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | RKHS-BA: A Robust Correspondence-Free Multi-View Registration Framework with Semantic Point Clouds Zhang, Ray Song, Jingwei Gao, Xiang Wu, Junzhe Liu, Tianyi Zhang, Jinyuan Eustice, Ryan Ghaffari, Maani Robotics This work reports a novel multi-frame Bundle Adjustment (BA) framework called RKHS-BA. It uses continuous landmark representations that encode RGB-D/LiDAR and semantic observations in a Reproducing Kernel Hilbert Space (RKHS). With a correspondence-free pose graph formulation, the proposed system constructs a loss function that achieves more generalized convergence than classical point-wise convergence. We demonstrate its applications in multi-view point cloud registration, sliding-window odometry, and global LiDAR mapping on simulated and real data. It shows highly robust pose estimations in extremely noisy scenes and exhibits strong generalization with various types of semantic inputs. The open source implementation is released in https://github.com/UMich-CURLY/RKHS_BA. |
| title | RKHS-BA: A Robust Correspondence-Free Multi-View Registration Framework with Semantic Point Clouds |
| topic | Robotics |
| url | https://arxiv.org/abs/2403.01254 |