Automated, objective texture segmentation of multibeam echosounder data

Describes use of multibeam echosounding (MBES) data collected by the Olympic Coast National Marine Sanctuary and a contracted seafloor survey, to illustrate an objective repeatable process that uses parameterized local Fourier histogram texture features to automate segmentation of surficial sediments from multiple types of acoustic imagery with a maximum likely decision rule. Also describes use of video from a towed camera sled which was integrated with sedimentary samples, backscatter, and bathymetry data to describe geological and biological aspects of habitat. Authors used a hierarchical deep-water marine benthic habitat classification scheme to create and attribute polygon features for use in a geographic information system. This report provides a description of the mapping and groundtruthing efforts, and technique and results of the automated segmentation procedure for each area surveyed in 2002 and 2003. The composition and texture of surficial sediments is recognized as being an important element of essential fish habitat, and the description and identification of EFH is a requirement of all fishery management plans under the Magnuson-Stevens Fisheries Conservation Act.