Data for Scallop (Pecten maximus) Identification in Natural Marine Habitats: A NetHarn Model Approach.
Authors/Creators
-
1.
University of the Highlands and Islands
-
2.
European Molecular Biology Laboratory
-
3.
University of St Andrews
Description
The data have been divided into distinct training and testing subsets, each encompassing still images corresponding to individual stations along with their respective annotations or predictions as CSV files.
This research explores the potential of Artificial Intelligence (AI), specifically the NetHarn model provided by the VIAME toolkit, to identify and count king and queen scallops from towed underwater video transects. The study utilizes video footage from NatureScot, captured using custom camera systems (DDV and miniDDV), providing a diverse dataset with variations in habitat, image quality, and camera specifications. Necessary details from the original report by Pascoe et al. (2021) are provided in the manuscript.
Pasco, G., James, B., Burke, L., Johnston, C., Orr, K., Clarke, J., Thorburn, J., Boulcott, P., Kent, F., Kamphausen, L. and Sinclair, R. (2021) 'Engaging the Fishing Industry in Marine Environmental Survey and Monitoring Scottish Marine and Freshwater Science Vol 12 No 3'. doi:10.7489/12365-1
Files
Test files.zip
Additional details
Related works
- Continues
- Peer review: 10.1016/j.ecoinf.2021.101233 (DOI)