BoxGraph: semantic place recognition and pose estimation from 3D LiDAR
This paper is about extremely robust and lightweight localisation using LiDAR point clouds based on instance segmentation and graph matching. We model 3D point clouds as fully-connected graphs of semantically identified components where each vertex corresponds to an object instance and encodes its s...
| Main Authors: | , , , |
|---|---|
| Format: | Conference item |
| Language: | English |
| Published: |
IEEE
2021
|
| _version_ | 1826309372116992000 |
|---|---|
| author | Pramatarov, G De Martini, D Gadd, M Newman, P |
| author_facet | Pramatarov, G De Martini, D Gadd, M Newman, P |
| author_sort | Pramatarov, G |
| collection | OXFORD |
| description | This paper is about extremely robust and lightweight localisation using LiDAR point clouds based on instance segmentation and graph matching. We model 3D point clouds as fully-connected graphs of semantically identified components where each vertex corresponds to an object instance and encodes its shape. Optimal vertex association across graphs allows for full 6-Degree-of-Freedom (DoF) pose estimation and place recognition by measuring similarity. This representation is very concise, condensing the size of maps by a factor of 25 against the state-of-the-art, requiring only 3 kB to represent a 1.4 MB laser scan. We verify the efficacy of our system on the SemanticKITTI dataset, where we achieve a new state-of-the-art in place recognition, with an average of 88.4 % recall at 100 % precision where the next closest competitor follows with 64.9 %. We also show accurate metric pose estimation performance - estimating 6-DoF pose with median errors of 10cm and 0.33 deg. |
| first_indexed | 2024-03-07T07:34:42Z |
| format | Conference item |
| id | oxford-uuid:49f84a11-fff9-4e27-a391-e8b05270af4b |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T07:34:42Z |
| publishDate | 2021 |
| publisher | IEEE |
| record_format | dspace |
| spelling | oxford-uuid:49f84a11-fff9-4e27-a391-e8b05270af4b2023-02-28T12:46:48ZBoxGraph: semantic place recognition and pose estimation from 3D LiDARConference itemhttp://purl.org/coar/resource_type/c_5794uuid:49f84a11-fff9-4e27-a391-e8b05270af4bEnglishSymplectic ElementsIEEE2021Pramatarov, GDe Martini, DGadd, MNewman, PThis paper is about extremely robust and lightweight localisation using LiDAR point clouds based on instance segmentation and graph matching. We model 3D point clouds as fully-connected graphs of semantically identified components where each vertex corresponds to an object instance and encodes its shape. Optimal vertex association across graphs allows for full 6-Degree-of-Freedom (DoF) pose estimation and place recognition by measuring similarity. This representation is very concise, condensing the size of maps by a factor of 25 against the state-of-the-art, requiring only 3 kB to represent a 1.4 MB laser scan. We verify the efficacy of our system on the SemanticKITTI dataset, where we achieve a new state-of-the-art in place recognition, with an average of 88.4 % recall at 100 % precision where the next closest competitor follows with 64.9 %. We also show accurate metric pose estimation performance - estimating 6-DoF pose with median errors of 10cm and 0.33 deg. |
| spellingShingle | Pramatarov, G De Martini, D Gadd, M Newman, P BoxGraph: semantic place recognition and pose estimation from 3D LiDAR |
| title | BoxGraph: semantic place recognition and pose estimation from 3D LiDAR |
| title_full | BoxGraph: semantic place recognition and pose estimation from 3D LiDAR |
| title_fullStr | BoxGraph: semantic place recognition and pose estimation from 3D LiDAR |
| title_full_unstemmed | BoxGraph: semantic place recognition and pose estimation from 3D LiDAR |
| title_short | BoxGraph: semantic place recognition and pose estimation from 3D LiDAR |
| title_sort | boxgraph semantic place recognition and pose estimation from 3d lidar |
| work_keys_str_mv | AT pramatarovg boxgraphsemanticplacerecognitionandposeestimationfrom3dlidar AT demartinid boxgraphsemanticplacerecognitionandposeestimationfrom3dlidar AT gaddm boxgraphsemanticplacerecognitionandposeestimationfrom3dlidar AT newmanp boxgraphsemanticplacerecognitionandposeestimationfrom3dlidar |