C. Tsukarnoto, I. Yuh, Autonomous Systems Laboratory Department of Mechanical EngineeringUniversity of Hawaii Honolulu, HI 96822Abstract
We describe a vision-based system for automatic ROV/AUV station keeping, based on an algorithmfor computing the motion of a camera (and the vehicle) from the images of the ocean bottom. The visionmodule estimates the vehicle position (displacement from station) and feeds this information to thecontrol system in order to generate the appropriate thruster signals. The current testing platform is aone-degree-of-freedom thruster system, submerged in a water tank and moving on a linear track. A PIDcontroller has been used without any attempt to optimize transient response performance by choosing thebest gains. The performance of the optical station keeping system is demonstrated in three selectedexperiments by comparing results with the data gathered from an optical encoder position sensor.
(2)Dynamic vision for ROY stabilization
S.Wasielewski, M.J. Alden
LIRMM 161 Rue Ada 34392 Montpelier cedex 5, FRANCE
Abstract -Stationkeeping is an important capability for various functions involving the operation ofundersea vehicles. This involves an accurate estimation of the vehicle displacement. ln this paper wepropose a monocular vision system for determining the motion of an underwater Remotely OperatedVehicle. The camera motion is estimated with a feature-based method which requires the extraction andthe matching of relevant features. The performance of the algorithms is presented with simulated andsynthetic images.
(3)Motion Estimation for ROV Stabilization with a Light-stripe Sensor
Jerome Vaganay and Brtmo Jouvencel, LJRMM 161, rue Ada 34392 Montpelier cedex 5, France
Abstract- Since the late 1970 s, the usefulness of Remotely Operated Vehicles (ROV) in fieldapplications has been largely demonstrated. For concerns such as operational cat, simplicity of use, andperformance, the trend is now at the automation of teleoperated tasks, so that autonomous ROV stabilization procedures are desired. This requires means of estimating the relative motion between theplatform and the offshore structure. This paper addresses the problem of relative motion estimation byconsidering an offshore structure made of cylinders and a light-stripe sensor. Simulation results of theproposed algorithm are presented. An experimental setup, currently tested 'in air, is then described. It isbased on a new generation of smart cameras, which integrate the light-sensitive elements and three imageprocessors on a single chip, and driven by a trausputer network.
