A GLOBAL LOCALIZATION SYSTEM FOR STRUCTURED ENVIRONMENTS USING AN IMAGING SONAR

This paper describes a localization system for Autonomous Underwater Vehicles (AUV) in structured underwater environments. It uses a Doppler Velocity Log (DVL) sensor, a pressure sensor, a compass and a mechanically scanning imaging sonar. An Extended Kalman Filter (EKF) merges the information from the DVL and the compass to estimate the robot trajectory. In order to reduce the drift inherent to this process we use acoustic images from an imaging sonar to determine the absolute robot position in the environment and update the estimated trajectory.