Automated image analysis for identifying the biofouling risk of vessels
CEBRA is developing a new image analysis tool to automatically identify biofouling of ship hulls from underwater photos which will make marine biosecurity inspections more accurate and efficient.
Biofouling — the accumulation of organisms on the surfaces of oceangoing vessels — is a major pathway for the introduction of non-indigenous marine species. Once introduced, they can cause severe social, environmental, and economic impacts. Current practice to manage biofouling risk uses costly, time consuming in-water inspections.
Developing an automated system to identify and classify biofouling risk on vessels will enable efficient assessment of biofouling on vessels. The system developed in this project will build on previous CEBRA work demonstrating the accuracy of image analysis to automatically detect the level of biofouling in images. New innovations aim to develop a more efficient methodology for labelling biofouling imagery datasets, analyse video data, and develop a practical user-interface for marine biosecurity agency staff.
CEBRA Project Leaders: Nathaniel Bloomfield, Andrew Robinson