Automating the assessment of biofouling in images using expert agreement as a gold standard

Nathaniel J Bloomfield,Peter Wilkinson,Susan Wei,Bartholomew A Woodham,Andrew P Robinson

doi:10.1038/s41598-021-81011-2

Nathaniel J Bloomfield, Peter Wilkinson + Show 3 more

Open Access

https://doi.org/10.1038/s41598-021-81011-2

Copy DOI

Abstract

Biofouling is the accumulation of organisms on surfaces immersed in water. It is of particular concern to the international shipping industry because it increases fuel costs and presents a biosecurity risk by providing a pathway for non-indigenous marine species to establish in new areas. There is growing interest within jurisdictions to strengthen biofouling risk-management regulations, but it is expensive to conduct in-water inspections and assess the collected data to determine the biofouling state of vessel hulls. Machine learning is well suited to tackle the latter challenge, and here we apply deep learning to automate the classification of images from in-water inspections to identify the presence and severity of fouling. We combined several datasets to obtain over 10,000 images collected from in-water surveys which were annotated by a group biofouling experts. We compared the annotations from three experts on a 120-sample subset of these images, and found that they showed 89% agreement (95% CI: 87–92%). Subsequent labelling of the whole dataset by one of these experts achieved similar levels of agreement with this group of experts, which we defined as performing at most 5% worse (p = 0.009–0.054). Using these expert labels, we were able to train a deep learning model that also agreed similarly with the group of experts (p = 0.001–0.014), demonstrating that automated analysis of biofouling in images is feasible and effective using this method.

Highlights

Biofouling is the accumulation of organisms on surfaces immersed in water
Global trade relies on the international shipping industry, which has been implicated in the spread of many marine non-indigenous species (NIS) around the w orld[1,2]
We assembled a dataset of 10,263 images collected from in-water surveys of around 300 commercial and recreational vessels. This dataset comprised images provided by three jurisdictions, namely: the Australian Department of Agriculture, Water and the Environment (DAWE), the New Zealand Ministry for Primary Industries (MPI), and the California State Lands Commission (CSLC)

Summary

Introduction

Biofouling is the accumulation of organisms on surfaces immersed in water. It is of particular concern to the international shipping industry because it increases fuel costs and presents a biosecurity risk by providing a pathway for non-indigenous marine species to establish in new areas. A vessel’s operating profile contributes to fouling risk, with extended periods of inactivity being associated with higher biofouling p ressure[18] Niche areas, such as sea chests, propellers, and other complex surface structures are at high risk of becoming fouled as they can offer a sheltered environment for fouling organisms to establish. For vessels staying longer than three weeks or visiting areas other than those designated as places of first arrival, any macrofouling except for goose barnacles is considered to be a biosecurity risk, while for short stay vessels there are macrofouling coverage thresholds[16,23] In implementing this policy, they have stressed the vessel management requirements rather than the thresholds, as even with current best management practices ships can become fouled[16]. Has proposed requirements for vessels to implement biofouling management practices or provide evidence that their fouling is appropriately controlled[21]

Objectives

Methods

Results

Conclusion