Title :
Real-time side scan image generation and registration framework for AUV route following
Author :
King, Peter ; Vardy, A. ; Vandrish, Peter ; Anstey, Benjamin
Author_Institution :
Marine Environ. Res. Lab. for Intell. Vehicles, Memorial Univ. of Newfoundland, St. John´´s, NL, Canada
Abstract :
Memorial University is in the development stages of a Qualitative Navigation System (QNS) to be deployed on the Memorial Explorer AUV. This system will allow localization and path following along a trained route without the necessity of a globally referenced position estimate. Previous QNS work has been on terrestrial robots using optical images. Our main challenge lies in utilization of side scan sonar as the imaging medium, as this type of sonar is prevalent on AUVs and provides much better range and coverage than optics in water. To achieve this, a sonar image processing and registration framework has been developed. To be useful such a framework should be fully-autonomous, robust, and operate in real-time, where real-time operation is defined as the ability to process, register and localize data at the rate it is collected, or faster. In this paper we describe our framework for processing sonar data, generating image tiles, extracting unique features and localizing against a reference set. We also present some results of this system based on raw sonar input data collected by the AUV.
Keywords :
autonomous underwater vehicles; feature extraction; image registration; mobile robots; path planning; position control; real-time systems; robot vision; sonar imaging; telerobotics; training; AUV route following; Memorial Explorer AUV; Memorial University; QNS; QNS work; data localization; data registeration; development stages; globally referenced position estimation; image processing; image tile generation; imaging medium; optical images; path following; qualitative navigation system; real-time operation; real-time side scan image generation; real-time side scan image registration; route training; side scan sonar utilization; sonar data processing; terrestrial robots; unique features extraction; Feature extraction; Real-time systems; Sonar; Sonar navigation; Tiles; Training;
Conference_Titel :
Autonomous Underwater Vehicles (AUV), 2012 IEEE/OES
Conference_Location :
Southampton
Print_ISBN :
978-1-4577-2055-0
Electronic_ISBN :
1522-3167
DOI :
10.1109/AUV.2012.6380758