A Streamlined and Standardized Benthic Habitat Mapping Approach for Marine and Hydrokinetic Site Environmental Assessments
We have combined two seafloor mapping techniques to efficiently map benthic habitat conditions. Multibeam echosounder (MBES) surveys provide detailed bathymetry and backscatter mosaics. Sediment profile and plan view imaging (SPI/PV) rapidly provides dense ground truth data on seabed physical and biological features. A key innovation of this project has been the development of a computer-assisted image processing system that standardizes the SPI data generation process. A unique combination of machine- and deep-learning techniques for grain-size classification, sediment–water interface delineation, and object detection (e.g., infauna and feeding voids) is described. An automated SPI image analysis system provides a standardized and cost-effective approach for image processing. Combining MBES with SPI/PV camera technology provides a first tier approach to rapidly documenting seabed physical and biological conditions in high resolution at MHK sites. When coupled with computer automated data extraction from the imagery, this approach provides a standardized and cost-effective method to establish and monitor benthic habitat conditions over time.