Automated underwater imaging system for continuous biofouling assessment

Traditional antifouling coating evaluations rely on manual sample retrieval, imaging and subjective visual assessments, which are labour-intensive and constrained by infrequent sampling intervals that may overlook critical fouling developments. This study presents a novel, automated biofouling assessment system at the CoaST Maritime Test Centre. A custom underwater camera captures daily images, while advanced image analysis techniques track fouling progression. To address imaging challenges, a median stacking technique removes transient artefacts such as overhanging algae and floating debris, enabling accurate assessment of attachment points, the primary metric for biofouling coverage. Multi-exposure fusion enhances image illumination. For image analysis, ilastik, a machine-learning tool using a random forest algorithm, performs pixel classification segmentation to differentiate biofouling classes and monitor their growth. This automated approach improves accuracy, repeatability and efficiency in biofouling assessment, reducing subjectivity and enhancing antifouling research.