Fast and accurate mapping of fine scale abundance of a VME in the deep sea with computer vision

Piechaud, Nils; Howell, Kerry L.

doi:10.1016/j.ecoinf.2022.101786

Cited by 12 publications

(7 citation statements)

References 75 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, imagery surveys are less invasive and can be conducted off a variety of platforms, making it more feasible to increase sampling effort and statistical power than trawl surveys. Current developments in seafloor imagery equipment (Clayton & Dennison, 2017;Dominguez-Carri o et al, 2021) and annotation (e.g., machine learning methods, Ayyagari et al, 2023;Piechaud & Howell, 2022), are expected to facilitate data collection and analysis. We acknowledge that our empirical data for sea pens are based on a single survey and do not capture temporal variation.…”

Section: Statistical Power Of Existing Data To Monitor Change In Co Taxamentioning

confidence: 99%

Monitoring data for a new large offshore marine protected area reveals infeasible management objectives

Morris,

Nguyen,

Neves

et al. 2024

Conservat Sci and Prac

View full text Add to dashboard Cite

Predicting and measuring changes resulting from marine protected areas (MPAs) has posed a challenge for practitioners, partly because ecosystems are complex and can change in unanticipated ways, but also due to MPA characteristics such as design factors, conservation objectives (COs), and monitoring programs, that can leave little chance of meeting stated goals. We consider these design factors for the Laurentian Channel MPA, a large offshore Canadian protected area established to protect against fishing impacts. Specifically, in this study we evaluated (1) whether it is realistic to expect improvements in the MPA for four previously established taxa‐specific COs, and (2) whether existing scientific surveys are capable of detecting changes in these CO taxa even if they occurred. Three CO species were sampled in scientific multispecies research vessel trawl surveys (Black Dogfish, Smooth Skate, and Northern Wolffish) and a fourth CO, sea pen taxa, were enumerated using seafloor imagery. Simulations indicate that trawl surveys have very little chance of detecting change in the abundance of the three fish species examined, while seafloor imagery data had higher statistical power for sea pen taxa. Moreover, we show that expecting change related to the removal of fishing is unrealistic due to the fact that the MPA was established in an area of minimal fishing pressure. While positive change is unlikely to be induced by the MPA, or be detected if they occurred, this MPA could provide conservation benefits if COs and monitoring approaches were realigned to match the unique features of this area that represents largely unimpacted sensitive benthic habitats.

show abstract

Section: Statistical Power Of Existing Data To Monitor Change In Co Taxamentioning

confidence: 99%

Monitoring data for a new large offshore marine protected area reveals infeasible management objectives

Morris,

Nguyen,

Neves

et al. 2024

Conservat Sci and Prac

View full text Add to dashboard Cite

show abstract

“…As YOLO does not possess annotation capabilities, the images used in training must have been previously labelled using a separate software. The capability of YOLO to automatically detect objects from marine images has been shown; YOLO version 4 [85] was used to develop a model to identify the Xenophyophore, Syringammina fragilissima (Brady, 1883), within 58 000 AUV video frames, requiring less than 10 days for complete analysis, and achieving a final precision of 0.91 and recall of 0.84 [86]. Additionally, the recently released YOLO version 8 [87] has been used to develop a model to simultaneously quantify the coral Dendrophyllia cornigera [44] and sponge Phakellia ventilabrum [44] within 5201 transect images [44].…”

Section: Machine Learning Tools For Marine Image Analysismentioning

confidence: 99%

New interactive machine learning tool for marine image analysis

Clark,

Smith,

McKay Fletcher

et al. 2024

R. Soc. Open Sci.

View full text Add to dashboard Cite

Advancing imaging technologies are drastically increasing the rate of marine video and image data collection. Often these datasets are not analysed to their full potential as extracting information for multiple species is incredibly time-consuming. This study demonstrates the capability of the open-source interactive machine learning tool, RootPainter , to analyse large marine image datasets quickly and accurately. The ability of RootPainter to extract the presence and surface area of the cold-water coral reef associate sponge species, Mycale lingua , was tested in two datasets: 18 346 time-lapse images and 1420 remotely operated vehicle video frames. New corrective annotation metrics integrated with RootPainter allow objective assessment of when to stop model training and reduce the need for manual model validation. Three highly accurate M. lingua models were created using RootPainter , with an average dice score of 0.94 ± 0.06. Transfer learning aided the production of two of the models, increasing analysis efficiency from 6 to 16 times faster than manual annotation for time-lapse images. Surface area measurements were extracted from both datasets allowing future investigation of sponge behaviours and distributions. Moving forward, interactive machine learning tools and model sharing could dramatically increase image analysis speeds, collaborative research and our understanding of spatiotemporal patterns in biodiversity.

show abstract

“…Moreover, unreliable classification schemes also prevents the replication of the classification approach on new datasets (Malik et al, 2023). Deep learning techniques offer an automated means to help mitigate this issue and reduce the time required/to analyse this data by orders of magnitude while providing a reliable measurement of the accuracy of the classification (Marrable et al, 2022;Piechaud and Howell, 2022).…”

Section: Object-based Image Analysismentioning

confidence: 99%

A review of new and existing non-extractive techniques for monitoring marine protected areas

et al. 2023

View full text Add to dashboard Cite

Ocean biodiversity loss is being driven by several anthropogenic threats and significant efforts are required to halt losses and promote healthy marine ecosystems. The establishment of a network of Marine Protected Areas (MPAs) can help restrict damaging activities and have been recognised as a potential solution to aid marine conservation. When managed correctly they can deliver both ecological and socio-economic benefits. In recent times, MPA designations have increased rapidly while many countries have set future MPA targets for the decades ahead. An integral element of MPA management is adequate monitoring that collects data to assess if conservation objectives are being achieved. Data acquired by monitoring can vary widely as can the techniques employed to collect such data. Ideally, non-destructive and non-invasive methods are preferred to prevent damage to habitats and species, though this may rule out a number of traditional extractive sampling approaches such as dredges and trawls. Moreover, advances in ocean observation technologies enable the collection of large amounts of data at high resolutions, while automated data processing is beginning to make analyses more logistically feasible and less time-consuming. Therefore, developments to existing marine monitoring techniques and new emerging technologies have led to a diverse array of options when choosing to implement an MPA monitoring programme. Here, we present a review of new and existing non-extractive techniques which can be applied to MPA monitoring. We summarise their capabilities, applications, advantages, limitations and possible future developments. The review is intended to aid MPA managers and researchers in determining the suitability of available monitoring techniques based on data requirements and site conditions.

show abstract

Fast and accurate mapping of fine scale abundance of a VME in the deep sea with computer vision

Cited by 12 publications

References 75 publications

Monitoring data for a new large offshore marine protected area reveals infeasible management objectives

Monitoring data for a new large offshore marine protected area reveals infeasible management objectives

New interactive machine learning tool for marine image analysis

A review of new and existing non-extractive techniques for monitoring marine protected areas

Contact Info

Product

Resources

About