Classification of seafloor habitats using acoustic backscattering of multibeam echosounder systems

Abstract

Recognition of the superior data and coverage from multibeam echosounder systems has been widespread and is increasing rapidly. These systems are currently the most advanced and efficient tool for remote observations and characterization of the seafloor. Modern, high resolution multibeam echosounder systems are capable of mapping acoustic backscattering strength coinciding with fine bathymetry, which improves substantially the capability of sonar to discriminate different types of seafloor habitats. Acoustic backscatter strength is highly dependent on incident angle due to different mechanisms of scattering within different angular domains. This study involves correction of the backscatter images to make them independent of incident angle using CMST MB Toolbox software to enable further segmentation and classification of the areas on the seafloor acoustically observed along multiple overlapping multibeam echosounder systems swath tracks. A set of bathymetry and acoustic backscattering data has been collected using 200 kHz RESON SeaBat 8124 multibeam echosounder. The ground truth data are also collected in a number of locations on the study area using grab samples of sediments to compare the unsupervised classification and statistical distribution of the backscattering energy attempts. The results were analyzed using error matrix statistical measures that provided a quantitative assessment of the classification results. Preliminary conclusions are made in this study with regard to the efficiency of seafloor classification using high resolution multibeam sonar systems.