Microbathymetry using self-contained navigation and simultaneous localization and mapping
Recent advances in high-resolution bathymetric sensors have entirely outpaced recent advances in underwater navigation. Producing microbathymetric maps therefore remains limited principally by navigation error. A method is proposed to fuse navigation data from a gyrocompass, Doppler velocity log, depth sensor, and a high-resolution bathymetric sensor to form a single self-consistent bathymetric map. The method combines both a standard Extended Kalman Filter (EKF) and the recently-developed incremental Smoothing and Mapping (iSAM) algorithm to produce maps with grid resolution under 5cm. Furthermore, the asymptotic runtime performance of iSAM's sparse square-root information smoothing approach allows production of maps that are large relative to the bathymetric sensor footprint. Finally, a method of hierarchically merging multiple maps is presented that overcomes the logistical limitations of surveys that often require 10's of hours of vehicle operations. The performance of the method is demonstrated on archaeological and geologic datasets from the 2010 and 2011 E/V Nautilus cruise seasons.^
Engineering, Marine and Ocean|Engineering, Robotics
James Ian S Vaughn,
"Microbathymetry using self-contained navigation and simultaneous localization and mapping"
Dissertations and Master's Theses (Campus Access).