Fast seafloor topography extraction and visualization from sparse AUV altimeter data

Most autonomous underwater vehicles (AUVs) are equipped to collect altimetry information, either from a simple altimeter for bottom avoidance, or as a product of a Doppler Velocity Log (DVL). This paper presents practical techniques for extracting and displaying coherent gridded maps of sea floor topography from such AUV altimeter data. In particular, methods for estimating and minimizing error in derived bathymetry data, as well as smoothing data for presentation and visualization are discussed. We discuss optimizing estimates of environmental parameters such as sound speed, as well as careful examination of lever arms. We present methods for interpolating between data points in order to fill data holidays, methods which are easily generalized to be used with many kinds of spatially distributed data. Although the density of data collected from DVL systems is sparse compared to full-scale multibeam sonars, this paper demonstrates that computed bathymetry from integrated altimeter data can be a key asset in interpreting other forms of AUV data, including side-scan sonar and oceanographic measurements. In part owing to the small size of the data sets, all of the algorithms presented in this paper are fast enough to be run without noticeably adding to post-processing time. We show examples of AUV operator tools that exploit fast gridded topography calculation in order to present data to the user.