On-water remote monitoring robotic system for estimating the patch coverage of Anabaena sp. filaments in shallow water

Abstract

An on-water remote monitoring robotic system was developed for indirectly estimating the relative density of marine cyanobacteria blooms at the subtidal sandy–rocky beach in Balandra Cove, Baja California Sur, Mexico. The system is based on an unmanned surface vehicle to gather underwater videos of the seafloor for avoiding physical damage on Anabaena sp. cyanobacteria colonies, which grow in tufts of filaments weakly attached to rocks, seagrass, and macroalgae. An on-axis image stabilization mechanism was developed to support a camcorder and minimize wave perturbation while recording underwater digital images of the seafloor. Color image processing algorithms were applied to estimate the patch coverage area and density, since Anabaena sp. filaments exhibit a characteristic green tone. Results of field tests showed the feasibility of the robotic system to estimate the relative density, distribution, and coverage area of cyanobacteria blooms, preventing the possible impact of direct observation. The robotic system could also be used in surveys of other benthos in the sublittoral zone.