G-Loc: Tightly-coupled Graph Localization with Prior Topo-metric Information
Montano-Oliván, Lorenzo, Placed, Julio A., Montano, Luis, Lázaro, María T.
–arXiv.org Artificial Intelligence
Localization in already mapped environments is a critical component in many robotics and automotive applications, where previously acquired information can be exploited along with sensor fusion to provide robust and accurate localization estimates. In this work, we offer a new perspective on map-based localization by reusing prior topological and metric information. Thus, we reformulate this long-studied problem to go beyond the mere use of metric maps. Our framework seamlessly integrates LiDAR, iner\-tial and GNSS measurements, and scan-to-map registrations in a sliding window graph fashion, which allows to accommodate the uncertainty of each observation. The modularity of our framework allows it to work with different sensor configurations (\textit{e.g.,} LiDAR resolutions, GNSS denial) and environmental conditions (\textit{e.g.,} map-less regions, large environments). We have conducted different validation experiments, including deployment in a real-world automotive application, demonstrating the accuracy, efficiency, and versatility of our system in online localization.
arXiv.org Artificial Intelligence
May-8-2024
- Country:
- North America > United States
- Georgia > Fulton County > Atlanta (0.04)
- Europe > Spain
- Aragón > Zaragoza Province > Zaragoza (0.04)
- Asia > Middle East
- Iran > Tehran Province > Tehran (0.04)
- North America > United States
- Genre:
- Research Report (0.64)
- Overview (0.46)
- Industry:
- Automobiles & Trucks (0.55)
- Transportation > Ground
- Road (0.47)
- Technology:
- Information Technology > Artificial Intelligence
- Vision (0.93)
- Machine Learning (0.93)
- Robots > Autonomous Vehicles (0.69)
- Representation & Reasoning > Optimization (0.46)
- Information Technology > Artificial Intelligence