DMCA
Real-time visual SLAM for autonomous underwater hull inspection using visual saliency (2013)
| Venue: | IEEE Transactions on Robotics |
| Citations: | 15 - 9 self |
BibTeX
@ARTICLE{Kim13real-timevisual,
author = {Ayoung Kim and Student Member and Ryan M. Eustice and Senior Member},
title = {Real-time visual SLAM for autonomous underwater hull inspection using visual saliency},
journal = {IEEE Transactions on Robotics},
year = {2013},
pages = {733}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Abstract—This paper reports on a real-time monocular visual simultaneous localization and mapping (SLAM) algorithm and results for its application in the area of autonomous underwater ship hull inspection. The proposed algorithm overcomes some of the specific challenges associated with underwater visual SLAM, namely limited field of view imagery and feature-poor regions. It does so by exploiting our SLAM navigation prior within the image registration pipeline and by being selective about which imagery is considered informative in terms of our visual SLAM map. A novel online bag-of-words measure for intra-and inter-image saliency are introduced, and are shown to be useful for image key-frame selection, information-gain based link hypothesis, and novelty detection. Results from three real-world hull inspection experiments evaluate the overall approach— including one survey comprising a 3.4 hour / 2.7 km long trajectory. Index Terms—SLAM, computer vision, marine robotics, visual saliency, information gain. I.
Keyphrases
visual saliency real-time visual slam autonomous underwater hull inspection intra-and inter-image saliency specific challenge feature-poor region underwater visual slam marine robotics image key-frame selection overall approach km long trajectory computer vision limited field real-world hull inspection experiment autonomous underwater ship hull inspection link hypothesis real-time monocular visual simultaneous localization visual slam map view imagery image registration pipeline index term slam slam navigation novel online bag-of-words measure information gain
