Change Detection of Sea Floor Environment Using Side Scan Sonar Data For Online Simultaneous Localization and Mapping on Autonomous Underwater Vehicles

Abstract

Autonomous underwater vehicles (AUVs) are frequently used to survey sea-floor environments using side-scan sonar technology. A simultaneous localization and mapping (SLAM) algorithm can be used with side-scan sonar data gathered during surveying to bound the possible error in AUV position estimate, and increase overall position accuracy, using only information already gathered during the survey mission. One problem in using SLAM to improve localization is that data from a preliminary or route survey on the sea floor may be inaccurate due to changes in the sea bed or merely be differently detected due to different side-scan sonar surveying patterns or equipment. This thesis’ focus is an integrated on-board SLAM system using automated target recognition system to extract objects for SLAM data association, data association algorithms for MLOs (joint compatibility program), and finally change detection on the SLAM results to determine if new objects have been introduced to the sea floor.