Show simple item record

dc.contributor.authorJackett, Christopher
dc.contributor.authorAlthaus, Franziska
dc.contributor.authorMaguire, Kylie
dc.contributor.authorFarazi, Moshiur
dc.contributor.authorScoulding, Ben
dc.contributor.authorUntiedt, Candice
dc.contributor.authorRyan, Tim
dc.contributor.authorShanks, Peter
dc.contributor.authorBrodie, Pamela
dc.contributor.authorWilliams, Alan
dc.date.accessioned2023-06-17T14:09:44Z
dc.date.available2023-06-17T14:09:44Z
dc.date.issued2023
dc.identifier.citationJackett, C., Althaus, F., Maguire, K., Farazi, M., Scoulding, B., Untiedt, C., Ryan, T., Shanks, P., Brodie, P.,and Williams, A. (2023) A benthic substrate classification method for seabed images using deep learning: Application to management of deep-sea coral reefs. Journal of Applied Ecology, 2023, 20pp. DOI: https://doi.org/10.1111/1365-2664.14408en_US
dc.identifier.urihttps://repository.oceanbestpractices.org/handle/11329/2291
dc.description.abstractProtecting deep-sea coral-based vulnerable marine ecosystems (VMEs) from human impacts, particularly bottom trawling, is a major conservation challenge in world oceans. Management processes for these ecosystems are weakened by key uncertainties that could be substantially addressed by having much greater volumes of quantitative image-derived data that detail the distribution and abundance of coral reefs and the nature of impacts upon them. Considerably greater volumes of data could be available if the resource costs of image annotation are reduced. In this paper we propose a solution: a deep learning system capable of automatically identifying reef-building stony corals amongst other seabed substrata in much larger volumes of seabed imagery than was previously possible. Using a previously annotated dataset, we trained a convolutional neural network on approximately 70,000 classified images (‘snips’) comprising six benthic substrate classes, including reef-building stony coral—‘coral matrix’. Model performance improvements, chiefly by dataset cleaning, transfer learning and hyperparameter optimisation, resulted in the final trained model achieving validation accuracy of 98.19%. The classification was robust: benthic substrate types were accurately differentiated, and in some cases more consistently than was achieved by human annotators. Synthesis and applications. The availability of much larger volumes of automatically annotated image-derived data will improve spatial management of impacts on coral-based VMEs in the deep sea by (1) improved cross-validation and performance of spatial models required to predict coral distribution and abundance over the large scales of managed areas, and (2) establishing empirical relationships between coral abundance on the seabed and coral bycatch landed during fishing operations.en_US
dc.language.isoenen_US
dc.rightsAttribution-NonCommercial 4.0 International*
dc.rights.urihttp://creativecommons.org/licenses/by-nc/4.0/*
dc.subject.otherBenthic substrateen_US
dc.subject.otherDeep learningen_US
dc.subject.otherUnderwater image analysisen_US
dc.subject.otherCoral matrixen_US
dc.subject.otherMachine learningen_US
dc.subject.otherSolenosmilia variabilisen_US
dc.subject.otherVulnerable marine ecosystemsen_US
dc.titleA benthic substrate classification method for seabed images using deep learning: Application to management of deep-sea coral reefs.en_US
dc.typeJournal Contributionen_US
dc.description.refereedRefereeden_US
dc.format.pagerange20pp.en_US
dc.identifier.doihttps://doi.org/10.1111/1365-2664.14408
dc.subject.parameterDisciplineOther biological measurementsen_US
dc.subject.parameterDisciplineUnderwater photographyen_US
dc.subject.dmProcessesData analysisen_US
dc.bibliographicCitation.titleJournal of Applied Ecologyen_US
dc.bibliographicCitation.volume2023en_US
dc.description.sdg14.aen_US
dc.description.maturitylevelPilot or Demonstrateden_US
dc.description.adoptionNovel (no adoption outside originators)en_US
dc.description.methodologyTypeSpecification of criteriaen_US
dc.description.methodologyTypeReports with methodological relevanceen_US
obps.contact.contactnameChris Jackett
obps.contact.contactemailchris.jackett@csiro.au
obps.resourceurl.publisherhttps://besjournals.onlinelibrary.wiley.com/doi/full/10.1111/1365-2664.14408


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

Attribution-NonCommercial 4.0 International
Except where otherwise noted, this item's license is described as Attribution-NonCommercial 4.0 International