Marine biofouling resistance rating using image analysis

Biofouling on ship hulls can cause increased fuel consumption and the global spread of nonindigenous species. The shipping industry utilizes fouling control coatings (FCC) to prevent biofouling from occurring on the ship hull. The evaluation standards for the performance of FCC are based on manual inspections, which unavoidably induce some degree of subjectivity. A standardized biofouling recognition model using image analysis would provide a more objective basis for the evaluation of FCC. For this purpose, several coated panels were immersed in the ocean at CoaST Maritime Test Centre for an exposure period of six weeks, whereafter, the panels were fully covered with biofouling. The program ilastik was then successfully used to train a pixel classification model, which could provide a simple segmentation of the different biofouling categories detected on a coated surface. From the simple segmentation, a coverage percentage of biofouling was determined. The percentages can stand alone to provide information on the degree of biofouling or be used in combination with the guidelines from the European Chemicals Agency to calculate a fouling resistance rating (FRR). The FRR obtained from the model was compared with FRR values obtained from the manual evaluation of the panels.

Automated summary

Biofouling on ship hulls can lead to increased fuel consumption and the spread of nonindigenous species. To objectively evaluate the performance of fouling control coatings (FCC), a standardized biofouling recognition model using image analysis was developed. This model provides a simple segmentation of different biofouling categories, allowing for the determination of biofouling coverage percentage. The results from the model were compared with manual evaluations to assess the accuracy of the model.