Home /Research /Real time visual SLAM for underwater robotic inspection
PERCEPTION

Real time visual SLAM for underwater robotic inspection

Magda Meireles, Renato Lourenço, André Dias, José Almeida, Hugo Silva, Alfredo Martins

Year
2014
Citations
14

Abstract

This paper addresses the development of an underwater visual navigation system for a Remotely Operated Vehicle (ROV) based on Real-Time Simultaneous Localization and Mapping method using natural landmarks. Our proposed approach was tested in an indoor tank, where field experiments were performed to obtain 3D vehicle (VIDEORAY Pro3 ROV) trajectory, and results validated using an external stereo vision “ground-truth” system.

Keywords

Remotely operated underwater vehicleSimultaneous localization and mappingComputer visionArtificial intelligenceUnderwaterComputer scienceTrajectoryGround truthMobile robotRobot

Related papers

Browse all PERCEPTION papers