.png){kind=icon}
.png){kind=icon}
Return
Map-based localization and loop-closure detection from a moving underwater platform using flow features
.png){kind=icon}
PRE
AI
DOI：10.1007/s10514-018-9797-3.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In recent years, flow sensing has gotten the attention of the robotics community as an exteroceptive sensing modality, in addition to the conventional underwater sensing modalities of vision and sonar. Earlier works on flow sensing for robotics focus on detection and characterization of objects' wakes, with the focus slowly evolving towards more complicated tasks such as localization of a stationary underwater platform using flow. In this paper we take this one step ahead, and present map-based localization and loop-closure detection from a continuously moving platform. Map-based localization is performed using flow features inside a particle filter framework, whereas loop-closure detection is based on indexation and comparison of flow features. Both techniques are validated by performing off-line experimentation on real flow data captured in complex flow inside a model fish pass. The results highlight the potential of using flow sensing (in addition to conventional underwater sensing modalities of vision and sonar) for the tasks of underwater robot perception and localization.
Keywords:
Flow sensing
Map-based localization
Loop-closure detection
Underwater robotics
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
4.3
Papers:
1.7K
Citations:
5.0K